research articles on teacher

• Our Mission

The 10 Most Significant Education Studies of 2021

From reframing our notion of “good” schools to mining the magic of expert teachers, here’s a curated list of must-read research from 2021.

It was a year of unprecedented hardship for teachers and school leaders. We pored through hundreds of studies to see if we could follow the trail of exactly what happened: The research revealed a complex portrait of a grueling year during which persistent issues of burnout and mental and physical health impacted millions of educators. Meanwhile, many of the old debates continued: Does paper beat digital? Is project-based learning as effective as direct instruction? How do you define what a “good” school is?

Other studies grabbed our attention, and in a few cases, made headlines. Researchers from the University of Chicago and Columbia University turned artificial intelligence loose on some 1,130 award-winning children’s books in search of invisible patterns of bias. (Spoiler alert: They found some.) Another study revealed why many parents are reluctant to support social and emotional learning in schools—and provided hints about how educators can flip the script.

1. What Parents Fear About SEL (and How to Change Their Minds)

When researchers at the Fordham Institute asked parents to rank phrases associated with social and emotional learning , nothing seemed to add up. The term “social-emotional learning” was very unpopular; parents wanted to steer their kids clear of it. But when the researchers added a simple clause, forming a new phrase—”social-emotional & academic learning”—the program shot all the way up to No. 2 in the rankings.

What gives?

Parents were picking up subtle cues in the list of SEL-related terms that irked or worried them, the researchers suggest. Phrases like “soft skills” and “growth mindset” felt “nebulous” and devoid of academic content. For some, the language felt suspiciously like “code for liberal indoctrination.”

But the study suggests that parents might need the simplest of reassurances to break through the political noise. Removing the jargon, focusing on productive phrases like “life skills,” and relentlessly connecting SEL to academic progress puts parents at ease—and seems to save social and emotional learning in the process.

2. The Secret Management Techniques of Expert Teachers

In the hands of experienced teachers, classroom management can seem almost invisible: Subtle techniques are quietly at work behind the scenes, with students falling into orderly routines and engaging in rigorous academic tasks almost as if by magic. 

That’s no accident, according to new research . While outbursts are inevitable in school settings, expert teachers seed their classrooms with proactive, relationship-building strategies that often prevent misbehavior before it erupts. They also approach discipline more holistically than their less-experienced counterparts, consistently reframing misbehavior in the broader context of how lessons can be more engaging, or how clearly they communicate expectations.

Focusing on the underlying dynamics of classroom behavior—and not on surface-level disruptions—means that expert teachers often look the other way at all the right times, too. Rather than rise to the bait of a minor breach in etiquette, a common mistake of new teachers, they tend to play the long game, asking questions about the origins of misbehavior, deftly navigating the terrain between discipline and student autonomy, and opting to confront misconduct privately when possible.

3. The Surprising Power of Pretesting

Asking students to take a practice test before they’ve even encountered the material may seem like a waste of time—after all, they’d just be guessing.

But new research concludes that the approach, called pretesting, is actually more effective than other typical study strategies. Surprisingly, pretesting even beat out taking practice tests after learning the material, a proven strategy endorsed by cognitive scientists and educators alike. In the study, students who took a practice test before learning the material outperformed their peers who studied more traditionally by 49 percent on a follow-up test, while outperforming students who took practice tests after studying the material by 27 percent.

The researchers hypothesize that the “generation of errors” was a key to the strategy’s success, spurring student curiosity and priming them to “search for the correct answers” when they finally explored the new material—and adding grist to a 2018 study that found that making educated guesses helped students connect background knowledge to new material.

Learning is more durable when students do the hard work of correcting misconceptions, the research suggests, reminding us yet again that being wrong is an important milestone on the road to being right.

4. Confronting an Old Myth About Immigrant Students

Immigrant students are sometimes portrayed as a costly expense to the education system, but new research is systematically dismantling that myth.

In a 2021 study , researchers analyzed over 1.3 million academic and birth records for students in Florida communities, and concluded that the presence of immigrant students actually has “a positive effect on the academic achievement of U.S.-born students,” raising test scores as the size of the immigrant school population increases. The benefits were especially powerful for low-income students.

While immigrants initially “face challenges in assimilation that may require additional school resources,” the researchers concluded, hard work and resilience may allow them to excel and thus “positively affect exposed U.S.-born students’ attitudes and behavior.” But according to teacher Larry Ferlazzo, the improvements might stem from the fact that having English language learners in classes improves pedagogy , pushing teachers to consider “issues like prior knowledge, scaffolding, and maximizing accessibility.”

5. A Fuller Picture of What a ‘Good’ School Is

It’s time to rethink our definition of what a “good school” is, researchers assert in a study published in late 2020.⁣ That’s because typical measures of school quality like test scores often provide an incomplete and misleading picture, the researchers found.

The study looked at over 150,000 ninth-grade students who attended Chicago public schools and concluded that emphasizing the social and emotional dimensions of learning—relationship-building, a sense of belonging, and resilience, for example—improves high school graduation and college matriculation rates for both high- and low-income students, beating out schools that focus primarily on improving test scores.⁣

“Schools that promote socio-emotional development actually have a really big positive impact on kids,” said lead researcher C. Kirabo Jackson in an interview with Edutopia . “And these impacts are particularly large for vulnerable student populations who don’t tend to do very well in the education system.”

The findings reinforce the importance of a holistic approach to measuring student progress, and are a reminder that schools—and teachers—can influence students in ways that are difficult to measure, and may only materialize well into the future.⁣

6. Teaching Is Learning

One of the best ways to learn a concept is to teach it to someone else. But do you actually have to step into the shoes of a teacher, or does the mere expectation of teaching do the trick?

In a 2021 study , researchers split students into two groups and gave them each a science passage about the Doppler effect—a phenomenon associated with sound and light waves that explains the gradual change in tone and pitch as a car races off into the distance, for example. One group studied the text as preparation for a test; the other was told that they’d be teaching the material to another student.

The researchers never carried out the second half of the activity—students read the passages but never taught the lesson. All of the participants were then tested on their factual recall of the Doppler effect, and their ability to draw deeper conclusions from the reading.

The upshot? Students who prepared to teach outperformed their counterparts in both duration and depth of learning, scoring 9 percent higher on factual recall a week after the lessons concluded, and 24 percent higher on their ability to make inferences. The research suggests that asking students to prepare to teach something—or encouraging them to think “could I teach this to someone else?”—can significantly alter their learning trajectories.

7. A Disturbing Strain of Bias in Kids’ Books

Some of the most popular and well-regarded children’s books—Caldecott and Newbery honorees among them—persistently depict Black, Asian, and Hispanic characters with lighter skin, according to new research .

Using artificial intelligence, researchers combed through 1,130 children’s books written in the last century, comparing two sets of diverse children’s books—one a collection of popular books that garnered major literary awards, the other favored by identity-based awards. The software analyzed data on skin tone, race, age, and gender.

Among the findings: While more characters with darker skin color begin to appear over time, the most popular books—those most frequently checked out of libraries and lining classroom bookshelves—continue to depict people of color in lighter skin tones. More insidiously, when adult characters are “moral or upstanding,” their skin color tends to appear lighter, the study’s lead author, Anjali Aduki,  told The 74 , with some books converting “Martin Luther King Jr.’s chocolate complexion to a light brown or beige.” Female characters, meanwhile, are often seen but not heard.

Cultural representations are a reflection of our values, the researchers conclude: “Inequality in representation, therefore, constitutes an explicit statement of inequality of value.”

8. The Never-Ending ‘Paper Versus Digital’ War

The argument goes like this: Digital screens turn reading into a cold and impersonal task; they’re good for information foraging, and not much more. “Real” books, meanwhile, have a heft and “tactility”  that make them intimate, enchanting—and irreplaceable.

But researchers have often found weak or equivocal evidence for the superiority of reading on paper. While a recent study concluded that paper books yielded better comprehension than e-books when many of the digital tools had been removed, the effect sizes were small. A 2021 meta-analysis further muddies the water: When digital and paper books are “mostly similar,” kids comprehend the print version more readily—but when enhancements like motion and sound “target the story content,” e-books generally have the edge.

Nostalgia is a force that every new technology must eventually confront. There’s plenty of evidence that writing with pen and paper encodes learning more deeply than typing. But new digital book formats come preloaded with powerful tools that allow readers to annotate, look up words, answer embedded questions, and share their thinking with other readers.

We may not be ready to admit it, but these are precisely the kinds of activities that drive deeper engagement, enhance comprehension, and leave us with a lasting memory of what we’ve read. The future of e-reading, despite the naysayers, remains promising.

9. New Research Makes a Powerful Case for PBL

Many classrooms today still look like they did 100 years ago, when students were preparing for factory jobs. But the world’s moved on: Modern careers demand a more sophisticated set of skills—collaboration, advanced problem-solving, and creativity, for example—and those can be difficult to teach in classrooms that rarely give students the time and space to develop those competencies.

Project-based learning (PBL) would seem like an ideal solution. But critics say PBL places too much responsibility on novice learners, ignoring the evidence about the effectiveness of direct instruction and ultimately undermining subject fluency. Advocates counter that student-centered learning and direct instruction can and should coexist in classrooms.

Now two new large-scale studies —encompassing over 6,000 students in 114 diverse schools across the nation—provide evidence that a well-structured, project-based approach boosts learning for a wide range of students.

In the studies, which were funded by Lucas Education Research, a sister division of Edutopia , elementary and high school students engaged in challenging projects that had them designing water systems for local farms, or creating toys using simple household objects to learn about gravity, friction, and force. Subsequent testing revealed notable learning gains—well above those experienced by students in traditional classrooms—and those gains seemed to raise all boats, persisting across socioeconomic class, race, and reading levels.

10. Tracking a Tumultuous Year for Teachers

The Covid-19 pandemic cast a long shadow over the lives of educators in 2021, according to a year’s worth of research.

The average teacher’s workload suddenly “spiked last spring,” wrote the Center for Reinventing Public Education in its January 2021 report, and then—in defiance of the laws of motion—simply never let up. By the fall, a RAND study recorded an astonishing shift in work habits: 24 percent of teachers reported that they were working 56 hours or more per week, compared to 5 percent pre-pandemic.

The vaccine was the promised land, but when it arrived nothing seemed to change. In an April 2021 survey  conducted four months after the first vaccine was administered in New York City, 92 percent of teachers said their jobs were more stressful than prior to the pandemic, up from 81 percent in an earlier survey.

It wasn’t just the length of the work days; a close look at the research reveals that the school system’s failure to adjust expectations was ruinous. It seemed to start with the obligations of hybrid teaching, which surfaced in Edutopia ’s coverage of overseas school reopenings. In June 2020, well before many U.S. schools reopened, we reported that hybrid teaching was an emerging problem internationally, and warned that if the “model is to work well for any period of time,” schools must “recognize and seek to reduce the workload for teachers.” Almost eight months later, a 2021 RAND study identified hybrid teaching as a primary source of teacher stress in the U.S., easily outpacing factors like the health of a high-risk loved one.

New and ever-increasing demands for tech solutions put teachers on a knife’s edge. In several important 2021 studies, researchers concluded that teachers were being pushed to adopt new technology without the “resources and equipment necessary for its correct didactic use.” Consequently, they were spending more than 20 hours a week adapting lessons for online use, and experiencing an unprecedented erosion of the boundaries between their work and home lives, leading to an unsustainable “always on” mentality. When it seemed like nothing more could be piled on—when all of the lights were blinking red—the federal government restarted standardized testing .

Change will be hard; many of the pathologies that exist in the system now predate the pandemic. But creating strict school policies that separate work from rest, eliminating the adoption of new tech tools without proper supports, distributing surveys regularly to gauge teacher well-being, and above all listening to educators to identify and confront emerging problems might be a good place to start, if the research can be believed.

• Utility Menu

• Get Involved
• News & Events

qualtrics survey

Teacher effectiveness.

Featured Project: Best Foot Forward

Can video technology improve the classroom observation process? Find out with the Best Foot Forward Project.

Featured Resource: SDP Educator Diagnostic

SDP collaborated with the state of Delaware to illuminate patterns related to teacher recruitment, impact, and retention.

Research consistently shows that teaching is the single most important school-based factor in a student’s academic growth. As such, the topic of effective teaching is at the forefront of CEPR’s research, which includes large national projects, like the National Center for Teacher Effectiveness, and program evaluations, like that of the Boston Teacher Residency. It is CEPR’s goal to have such analyses inform policy decisions and aid in drawing implications for reform.

Other Focus Areas

• Postsecondary Access & Success
• School Improvement & Redesign
• Related Projects
• Presentations

Mary Beisiegel

Elizabeth Cascio

Dan Goldhaber

Heather C. Hill

... Read more about Heather C. Hill

... Read more about Andrew Ho

Thomas J. Kane

... Read more about Thomas J. Kane

Matthew Kraft

Bill Penuel

... Read more about Bill Penuel

Jonah Rockoff

While research has generated substantial information regarding the characteristics of effective mathematics teachers and classrooms, scholars have rarely tested multiple aspects of teachers or teaching within a single study. Without testing multiple variables simultaneously, it is difficult to identify specific aspects of mathematics teachers and teaching that may be particularly impactful on student learning, and to understand the degree to which these characteristics are related to one another. This plenary draws on data from a three-year study measuring multiple components of teacher and teaching quality to investigate these issues.

Best Foot Forward Project

Did video technology improve the classroom observation process?

The Best Foot Forward Project investigated whether video technology can make the classroom observation process easier to implement, less costly, and more valid and reliable. In a randomized controlled trial, the study team put cameras in the hands of teachers and allowed them to select their best lessons for evaluation. Researchers aimed to learn whether digital video made the observation process more acceptable to teachers and administrators.... Read more about Best Foot Forward Project

Developing Common Core Classrooms Through Rubric-Based Coaching

What is the impact of rubric-based coaching on teacher instructional practice and student achievement outcomes?

This research study included the development and implementation of the MQI Coaching Cycle, as well as a rigorous evaluation of the program’s effectiveness using a teacher-level random assignment design.... Read more about Developing Common Core Classrooms Through Rubric-Based Coaching

Exploring Methods for Improving Teachers’ Mathematical Quality of Instruction

What is the optimal level of facilitation in teachers’ discussions and the effect of  contributing their own video ?

This study is investigating the conditions of effective video-based math professional development and the use of an observational instrument as a common language for teachers to use to talk about their practice.... Read more about Exploring Methods for Improving Teachers’ Mathematical Quality of Instruction

Mathematical Knowledge for Teaching Measures: Refreshing the Item Pool

How do we measure the specialized content knowledge required to teacher mathematics?

The Mathematical Knowledge for Teaching (MKT) instrument is a tool for measuring the specialized knowledge K-8 teachers of mathematics use in teaching. Since its creation, MKT items have been widely used in evaluating teacher learning in professional development programs and investigating the relationship between teachers’ knowledge and practice.... Read more about Mathematical Knowledge for Teaching Measures: Refreshing the Item Pool

Mathematical Quality of Instruction (MQI)

The MQI is a Common Core-aligned observational rubric that provides a framework for analyzing mathematics instruction in several domains. Within each of the five domains, individual codes contain score points that categorize instruction into different levels of quality. The MQI was developed in order to provide a both multidimensional and balanced view of mathematics instruction.... Read more about Mathematical Quality of Instruction (MQI)

MET Project

CEPR Faculty Director Tom Kane led the the Measures of Effective Teaching (MET) project, a $52 million study sponsored by the Bill & Melinda Gates Foundation. Dramatically better student outcomes will require dramatically different teaching. And dramatically different teaching will require better feedback for teachers. The project brought together 3,000 teacher volunteers in six different school districts with dozens of education experts and researchers to reinvent the way teacher evaluations are done. 1

Middle School Mathematics Teachers and Teaching Survey

What is the current state of mathematics education in the United States?

We suspect that mathematics teaching has changed tremendously in the last two decades, as new teaching methods, technologies and curriculum materials have appeared in classrooms. This study has been designed to understand the consequences of these changes for instruction.... Read more about Middle School Mathematics Teachers and Teaching Survey

National Center for Teacher Effectiveness

How are multiple measures used in teacher evaluation related to one another and student learning? 

In July 2009, NCTE commenced a six-year effort to join disparate strands of education research, and develop a deeper and more comprehensive understanding of how to measure teacher and teaching effectiveness.... Read more about National Center for Teacher Effectiveness

The project team is still awaiting student test data to complete the evaluation, but this brief provides a short update on survey results. Students of MQI-coached teachers report that their teachers ask more substantive questions, and require more use of mathematical vocabulary as compared to students of control teachers. Students in MQI-coached classrooms also reported more student talk in class. Teachers who received MQI Coaching tended to find their professional development significantly more useful than control teachers, and were also more likely to report that their mathematics instruction improved over the course of the year.

(2016). Findings from a National Study on Research Use Among School and District Leaders (Technical Report No. 1) . National Center for Research in Policy and Practice. Read the report (NCRPP Website)

Recent investigations into the education production function have moved beyond traditional teacher inputs, such as education, certification, and salary, focusing instead on observational measures of teaching practice. However, challenges to identification mean that this work has yet to coalesce around specific instructional dimensions that increase student achievement. I build on this discussion by exploiting within-school, between-grade, and cross-cohort variation in scores from two observation instruments; further, I condition on a uniquely rich set of teacher characteristics, practices, and skills. Findings indicate that inquiry-oriented instruction positively predicts student achievement. Content errors and imprecisions are negatively related, though these estimates are sensitive to the set of covariates included in the model. Two other dimensions of instruction, classroom emotional support and classroom organization, are not related to this outcome. Findings can inform recruitment and development efforts aimed at improving the quality of the teacher workforce. 

This document provides information on the grades 4 and 5 mathematics test modules developed as a joint Harvard/ETS venture from 2009 to 2012. Hundreds of items that were initially developed were then culled by a series of reviews, revisions, pilot tests, cognitive interviews, and field tests. 

The SDP Toolkit for Effective Data Use is a resource guide for education agency analysts who collect and analyze data on student achievement. Completing the toolkit produces a set of basic, yet essential, human capital and college-going analyses that every education agency should have as a foundation to inform strategic management and policy decisions.

An important assumption underlying meaningful comparisons of scores in rater-mediated assessments is that measurement is commensurate across raters. When raters differentially apply the standards established by an instrument, scores from different raters are on fundamentally different scales and no longer preserve a common meaning and basis for comparison. In this study, we developed a method to accommodate measurement noninvariance across raters when measurements are cross-classified within two distinct hierarchical units. We conceptualized random item effects cross-classified graded response models and used random discrimination and threshold effects to test, calibrate, and account for measurement noninvariance among raters. By leveraging empirical estimates of rater-specific deviations in the discrimination and threshold parameters, the proposed method allows us to identify noninvariant items and empirically estimate and directly adjust for this noninvariance within a cross-classified framework. Within the context of teaching evaluations, the results of a case study suggested substantial noninvariance across raters and that establishing an approximately invariant scale through random item effects improves model fit and predictive validity.

Education Decision Makers Need More Timely, Actionable Data

CEPR Executive Director Jon Fullerton and Proving Ground Director Bi Vuong discuss the need to utilize short-term intermediate measures of success to effectively manage interventions within school districts in the Stanford Social Innovation Review. ... Read more about Education Decision Makers Need More Timely, Actionable Data

How to Get Better

The Best Foot Forward Toolkit, as well as its potential for helping teachers become better educators, is referenced in the following HGSE Usable Knowledge blog post.... Read more about How to Get Better

Partnering in Education Research Program Accepting Applications

Learn more about the Partnering in Education Research (PIER) Program from the following HGSE article. ... Read more about Partnering in Education Research Program Accepting Applications

Using video in the classroom to help teachers and students

The Best Foot Forward project is highlighted in the following Education Dive article about video observations in the classroom. ... Read more about Using video in the classroom to help teachers and students

The Benefits of Adding Video to Teacher Evaluations

CEPR Director of Education and Communications, Miriam Greenberg, shares the benefits of video technology to teacher feedback in the following the following eSchool News article. ... Read more about The Benefits of Adding Video to Teacher Evaluations

Harvard Center Releases Report on First Full Year of Common Core Implementation

Research Shows Teachers Embraced New Standards and Adjusted Instruction to Meet Higher Bar; Provides New Evidence on Strategies Related to Improved Student Performance.... Read more about Harvard Center Releases Report on First Full Year of Common Core Implementation

A Toolkit for Video in the Classroom How Classroom-Based, Teacher-Controlled Video Can Improve the Observation Process

The Best Foot Forward Project's newly released toolkit is featured in the following Usable Knowledge blog post.... Read more about A Toolkit for Video in the Classroom How Classroom-Based, Teacher-Controlled Video Can Improve the Observation Process

Report: Teacher-Controlled Video Observations Improve Teacher Assessment Process

The Best Foot Forward Project's report on video observations for teachers is featured in the following The Journal article.... Read more about Report: Teacher-Controlled Video Observations Improve Teacher Assessment Process

Let’s Rewind

Miriam Greenberg, Director of Education and Communications, shares the importance of video technology to teacher feedback in the following HGSE Usable Knowledge blog post. ... Read more about Let’s Rewind

Harvard Center’s Best Foot Forward Project Shares Results on the Use of Video in Classroom Observations

Research shows the use of video in teacher observations is an effective tool for improving evaluation feedback. majority of teachers found process fairer and more useful overall.... read more about harvard center’s best foot forward project shares results on the use of video in classroom observations.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• HHS Author Manuscripts

Teacher and Teaching Effects on Students’ Attitudes and Behaviors

David blazar.

Harvard Graduate School of Education

Matthew A. Kraft

Brown University

Associated Data

Research has focused predominantly on how teachers affect students’ achievement on tests despite evidence that a broad range of attitudes and behaviors are equally important to their long-term success. We find that upper-elementary teachers have large effects on self-reported measures of students’ self-efficacy in math, and happiness and behavior in class. Students’ attitudes and behaviors are predicted by teaching practices most proximal to these measures, including teachers’ emotional support and classroom organization. However, teachers who are effective at improving test scores often are not equally effective at improving students’ attitudes and behaviors. These findings lend empirical evidence to well-established theory on the multidimensional nature of teaching and the need to identify strategies for improving the full range of teachers’ skills.

1. Introduction

Empirical research on the education production function traditionally has examined how teachers and their background characteristics contribute to students’ performance on standardized tests ( Hanushek & Rivkin, 2010 ; Todd & Wolpin, 2003 ). However, a substantial body of evidence indicates that student learning is multidimensional, with many factors beyond their core academic knowledge as important contributors to both short- and long-term success. 1 For example, psychologists find that emotion and personality influence the quality of one’s thinking ( Baron, 1982 ) and how much a child learns in school ( Duckworth, Quinn, & Tsukayama, 2012 ). Longitudinal studies document the strong predictive power of measures of childhood self-control, emotional stability, persistence, and motivation on health and labor market outcomes in adulthood ( Borghans, Duckworth, Heckman, & Ter Weel, 2008 ; Chetty et al., 2011 ; Moffitt et. al., 2011 ). In fact, these sorts of attitudes and behaviors are stronger predictors of some long-term outcomes than test scores ( Chetty et al., 2011 ).

Consistent with these findings, decades worth of theory also have characterized teaching as multidimensional. High-quality teachers are thought and expected not only to raise test scores but also to provide emotionally supportive environments that contribute to students’ social and emotional development, manage classroom behaviors, deliver accurate content, and support critical thinking ( Cohen, 2011 ; Lampert, 2001 ; Pianta & Hamre, 2009 ). In recent years, two research traditions have emerged to test this theory using empirical evidence. The first tradition has focused on observations of classrooms as a means of identifying unique domains of teaching practice ( Blazar, Braslow, Charalambous, & Hill, 2015 ; Hamre et al., 2013 ). Several of these domains, including teachers’ interactions with students, classroom organization, and emphasis on critical thinking within specific content areas, aim to support students’ development in areas beyond their core academic skill. The second research tradition has focused on estimating teachers’ contribution to student outcomes, often referred to as “teacher effects” ( Chetty Friedman, & Rockoff, 2014 ; Hanushek & Rivkin, 2010 ). These studies have found that, as with test scores, teachers vary considerably in their ability to impact students’ social and emotional development and a variety of observed school behaviors ( Backes & Hansen, 2015 ; Gershenson, 2016 ; Jackson, 2012 ; Jennings & DiPrete, 2010 ; Koedel, 2008 ; Kraft & Grace, 2016 ; Ladd & Sorensen, 2015 ; Ruzek et al., 2015 ). Further, weak to moderate correlations between teacher effects on different student outcomes suggest that test scores alone cannot identify teachers’ overall skill in the classroom.

Our study is among the first to integrate these two research traditions, which largely have developed in isolation. Working at the intersection of these traditions, we aim both to minimize threats to internal validity and to open up the “black box” of teacher effects by examining whether certain dimensions of teaching practice predict students’ attitudes and behaviors. We refer to these relationships between teaching practice and student outcomes as “teaching effects.” Specifically, we ask three research questions:

• To what extent do teachers impact students’ attitudes and behaviors in class?
• To what extent do specific teaching practices impact students’ attitudes and behaviors in class?
• Are teachers who are effective at raising test-score outcomes equally effective at developing positive attitudes and behaviors in class?

To answer our research questions, we draw on a rich dataset from the National Center for Teacher Effectiveness of upper-elementary classrooms that collected teacher-student links, observations of teaching practice scored on two established instruments, students’ math performance on both high- and low-stakes tests, and a student survey that captured their attitudes and behaviors in class. We used this survey to construct our three primary outcomes: students’ self-reported self-efficacy in math, happiness in class, and behavior in class. All three measures are important outcomes of interest to researchers, policymakers, and parents ( Borghans et al., 2008 ; Chetty et al., 2011 ; Farrington et al., 2012 ). They also align with theories linking teachers and teaching practice to outcomes beyond students’ core academic skills ( Bandura, Barbaranelli, Caprara, & Pastorelli, 1996 ; Pianta & Hamre, 2009 ), allowing us to test these theories explicitly.

We find that upper-elementary teachers have substantive impacts on students’ self-reported attitudes and behaviors in addition to their math performance. We estimate that the variation in teacher effects on students’ self-efficacy in math and behavior in class is of similar magnitude to the variation in teacher effects on math test scores. The variation of teacher effects on students’ happiness in class is even larger. Further, these outcomes are predicted by teaching practices most proximal to these measures, thus aligning with theory and providing important face and construct validity to these measures. Specifically, teachers’ emotional support for students is related both to their self-efficacy in math and happiness in class. Teachers’ classroom organization predicts students’ reports of their own behavior in class. Errors in teachers’ presentation of mathematical content are negatively related to students’ self-efficacy in math and happiness in class, as well as students’ math performance. Finally, we find that teachers are not equally effective at improving all outcomes. Compared to a correlation of 0.64 between teacher effects on our two math achievement tests, the strongest correlation between teacher effects on students’ math achievement and effects on their attitudes or behaviors is 0.19.

Together, these findings add further evidence for the multidimensional nature of teaching and, thus, the need for researchers, policymakers, and practitioners to identify strategies for improving these skills. In our conclusion, we discuss several ways that policymakers and practitioners may start to do so, including through the design and implementation of teacher evaluation systems, professional development, recruitment, and strategic teacher assignments.

2. Review of Related Research

Theories of teaching and learning have long emphasized the important role teachers play in supporting students’ development in areas beyond their core academic skill. For example, in their conceptualization of high-quality teaching, Pianta and Hamre (2009) describe a set of emotional supports and organizational techniques that are equally important to learners as teachers’ instructional methods. They posit that, by providing “emotional support and a predictable, consistent, and safe environment” (p. 113), teachers can help students become more self-reliant, motivated to learn, and willing to take risks. Further, by modeling strong organizational and management structures, teachers can help build students’ own ability to self-regulate. Content-specific views of teaching also highlight the importance of teacher behaviors that develop students’ attitudes and behaviors in ways that may not directly impact test scores. In mathematics, researchers and professional organizations have advocated for teaching practices that emphasize critical thinking and problem solving around authentic tasks ( Lampert, 2001 ; National Council of Teachers of Mathematics [NCTM], 1989 , 2014 ). Others have pointed to teachers’ important role of developing students’ self-efficacy and decreasing their anxiety in math ( Bandura et al., 1996 ; Usher & Pajares, 2008 ; Wigfield & Meece, 1988 ).

In recent years, development and use of observation instruments that capture the quality of teachers’ instruction have provided a unique opportunity to examine these theories empirically. One instrument in particular, the Classroom Assessment Scoring System (CLASS), is organized around “meaningful patterns of [teacher] behavior…tied to underlying developmental processes [in students]” ( Pianta & Hamre, 2009 , p. 112). Factor analyses of data collected by this instrument have identified several unique aspects of teachers’ instruction: teachers’ social and emotional interactions with students, their ability to organize and manage the classroom environment, and their instructional supports in the delivery of content ( Hafen et al., 2015 ; Hamre et al., 2013 ). A number of studies from developers of the CLASS instrument and their colleagues have described relationships between these dimensions and closely related student attitudes and behaviors. For example, teachers’ interactions with students predicts students’ social competence, engagement, and risk-taking; teachers’ classroom organization predicts students’ engagement and behavior in class ( Burchinal et al., 2008 ; Downer, Rimm-Kaufman, & Pianta, 2007 ; Hamre, Hatfield, Pianta, & Jamil, 2014 ; Hamre & Pianta, 2001 ; Luckner & Pianta, 2011 ; Mashburn et al., 2008 ; Pianta, La Paro, Payne, Cox, & Bradley, 2002 ). With only a few exceptions (see Downer et al., 2007 ; Hamre & Pianta, 2001 ; Luckner & Pianta, 2011 ), though, these studies have focused on pre-kindergarten settings.

Additional content-specific observation instruments highlight several other teaching competencies with links to students’ attitudes and behaviors. For example, in this study we draw on the Mathematical Quality of Instruction (MQI) to capture math-specific dimensions of teachers’ classroom practice. Factor analyses of data captured both by this instrument and the CLASS identified two teaching skills in addition to those described above: the cognitive demand of math activities that teachers provide to students and the precision with which they deliver this content ( Blazar et al., 2015 ). Validity evidence for the MQI has focused on the relationship between these teaching practices and students’ math test scores ( Blazar, 2015 ; Kane & Staiger, 2012 ), which makes sense given the theoretical link between teachers’ content knowledge, delivery of this content, and students’ own understanding ( Hill et al., 2008 ). However, professional organizations and researchers also describe theoretical links between the sorts of teaching practices captured on the MQI and student outcomes beyond test scores ( Bandura et al., 1996 ; Lampert, 2001 ; NCTM, 1989 , 2014 ; Usher & Pajares, 2008 ; Wigfield & Meece, 1988 ) that, to our knowledge, have not been tested.

In a separate line of research, several recent studies have borrowed from the literature on teachers’ “value-added” to student test scores in order to document the magnitude of teacher effects on a range of other outcomes. These studies attempt to isolate the unique effect of teachers on non-tested outcomes from factors outside of teachers’ control (e.g., students’ prior achievement, race, gender, socioeconomic status) and to limit any bias due to non-random sorting. Jennings and DiPrete (2010) estimated the role that teachers play in developing kindergarten and first-grade students’ social and behavioral outcomes. They found within-school teacher effects on social and behavioral outcomes that were even larger (0.21 standard deviations [sd]) than effects on students’ academic achievement (between 0.12 sd and 0.15 sd, depending on grade level and subject area). In a study of 35 middle school math teachers, Ruzek et al. (2015) found small but meaningful teacher effects on students’ motivation between 0.03 sd and 0.08 sd among seventh graders. Kraft and Grace (2016) found teacher effects on students’ self-reported measures of grit, growth mindset and effort in class ranging between 0.14 and 0.17 sd. Additional studies identified teacher effects on students’ observed school behaviors, including absences, suspensions, grades, grade progression, and graduation ( Backes & Hansen, 2015 ; Gershenson, 2016 ; Jackson, 2012 ; Koedel, 2008 ; Ladd & Sorensen, 2015 ).

To date, evidence is mixed on the extent to which teachers who improve test scores also improve other outcomes. Four of the studies described above found weak relationships between teacher effects on students’ academic performance and effects on other outcome measures. Compared to a correlation of 0.42 between teacher effects on math versus reading achievement, Jennings and DiPrete (2010) found correlations of 0.15 between teacher effects on students’ social and behavioral outcomes and effects on either math or reading achievement. Kraft and Grace (2016) found correlations between teacher effects on achievement outcomes and multiple social-emotional competencies were sometimes non-existent and never greater than 0.23. Similarly, Gershenson (2016) and Jackson (2012) found weak or null relationships between teacher effects on students’ academic performance and effects on observed schools behaviors. However, correlations from two other studies were larger. Ruzek et al. (2015) estimated a correlation of 0.50 between teacher effects on achievement versus effects on students’ motivation in math class. Mihaly, McCaffrey, Staiger, and Lockwood (2013) found a correlation of 0.57 between middle school teacher effects on students’ self-reported effort versus effects on math test scores.

Our analyses extend this body of research by estimating teacher effects on additional attitudes and behaviors captured by students in upper-elementary grades. Our data offer the unique combination of a moderately sized sample of teachers and students with lagged survey measures. We also utilize similar econometric approaches to test the relationship between teaching practice and these same attitudes and behaviors. These analyses allow us to examine the face validity of our teacher effect estimates and the extent to which they align with theory.

3. Data and Sample

Beginning in the 2010–2011 school year, the National Center for Teacher Effectiveness (NCTE) engaged in a three-year data collection process. Data came from participating fourth-and fifth-grade teachers (N = 310) in four anonymous, medium to large school districts on the East coast of the United States who agreed to have their classes videotaped, complete a teacher questionnaire, and help collect a set of student outcomes. Teachers were clustered within 52 schools, with an average of six teachers per school. While NCTE focused on teachers’ math instruction, participants were generalists who taught all subject areas. This is important, as it allowed us to isolate the contribution of individual teachers to students’ attitudes and behaviors, which is considerably more challenging when students are taught by multiple teachers. It also suggests that the observation measures, which assessed teachers’ instruction during math lessons, are likely to capture aspects of their classroom practice that are common across content areas.

In Table 1 , we present descriptive statistics on participating teachers and their students. We do so for the full NCTE sample, as well as for a subsample of teachers whose students were in the project in both the current and prior years. This latter sample allowed us to capture prior measures of students’ attitudes and behaviors, a strategy that we use to increase internal validity and that we discuss in more detail below. 2 When we compare these samples, we find that teachers look relatively similar with no statistically significant differences on any observable characteristic. Reflecting national patterns, the vast majority of elementary teachers in our sample are white females who earned their teaching credential through traditional certification programs. (See Hill, Blazar, & Lynch, 2015 for a discussion of how these teacher characteristics were measured.)

Participant Demographics

Students in our samples look similar to those in many urban districts in the United States, where roughly 68% are eligible for free or reduced-price lunch, 14% are classified as in need of special education services, and 16% are identified as limited English proficient; roughly 31% are African American, 39% are Hispanic, and 28% are white ( Council of the Great City Schools, 2013 ). We do observe some statistically significant differences between student characteristics in the full sample versus our analytic subsample. For example, the percentage of students identified as limited English proficient was 20% in the full sample compared to 14% in the sample of students who ever were part of analyses drawing on our survey measures. Although variation in samples could result in dissimilar estimates across models, the overall character of our findings is unlikely to be driven by these modest differences.

3.1. Students’ Attitudes and Behaviors

As part of the expansive data collection effort, researchers administered a student survey with items (N = 18) that were adapted from other large-scale surveys including the TRIPOD, the MET project, the National Assessment of Educational Progress (NAEP), and the Trends in International Mathematics and Science Study (TIMSS) (see Appendix Table 1 for a full list of items). Items were selected based on a review of the research literature and identification of constructs thought most likely to be influenced by upper-elementary teachers. Students rated all items on a five-point Likert scale where 1 = Totally Untrue and 5 = Totally True.

We identified a parsimonious set of three outcome measures based on a combination of theory and exploratory factor analyses (see Appendix Table 1 ). 3 The first outcome, which we call Self-Efficacy in Math (10 items), is a variation on well-known constructs related to students’ effort, initiative, and perception that they can complete tasks. The second related outcome measure is Happiness in Class (5 items), which was collected in the second and third years of the study. Exploratory factor analyses suggested that these items clustered together with those from Self-Efficacy in Math to form a single construct. However, post-hoc review of these items against the psychology literature from which they were derived suggests that they can be divided into a separate domain. As above, this measure is a school-specific version of well-known scales that capture students’ affect and enjoyment ( Diener, 2000 ). Both Self-Efficacy in Math and Happiness in Class have relatively high internal consistency reliabilities (0.76 and 0.82, respectively) that are similar to those of self-reported attitudes and behaviors explored in other studies ( Duckworth et al., 2007 ; John & Srivastava, 1999 ; Tsukayama et al., 2013 ). Further, self-reported measures of similar constructs have been linked to long-term outcomes, including academic engagement and earnings in adulthood, even conditioning on cognitive ability ( King, McInerney, Ganotice, & Villarosa, 2015 ; Lyubomirsky, King, & Diener, 2005 ).

The third and final construct consists of three items that were meant to hold together and which we call Behavior in Class (internal consistency reliability is 0.74). Higher scores reflect better, less disruptive behavior. Teacher reports of students’ classroom behavior have been found to relate to antisocial behaviors in adolescence, criminal behavior in adulthood, and earnings ( Chetty et al., 2011 ; Segal, 2013 ; Moffitt et al., 2011 ; Tremblay et al., 1992 ). Our analysis differs from these other studies in the self-reported nature of the behavior outcome. That said, other studies also drawing on elementary school students found correlations between self-reported and either parent- or teacher-reported measures of behavior that were similar in magnitude to correlations between parent and teacher reports of student behavior ( Achenbach, McConaughy, & Howell, 1987 ; Goodman, 2001 ). Further, other studies have found correlations between teacher-reported behavior of elementary school students and either reading or math achievement ( r = 0.22 to 0.28; Miles & Stipek, 2006 ; Tremblay et al., 1992 ) similar to the correlation we find between students’ self-reported Behavior in Class and our two math test scores ( r = 0.24 and 0.26; see Table 2 ). Together, this evidence provides both convergent and consequential validity evidence for this outcome measure. For all three of these outcomes, we created final scales by reverse coding items with negative valence and averaging raw student responses across all available items. 4 We standardized these final scores within years, given that, for some measures, the set of survey items varied across years.

Descriptive Statistics for Students' Academic Performance, Attitudes, and Behaviors

For high-stakes math test, reliability varies by district; thus, we report the lower bound of these estimates. Self-Efficacy in Math, Happiness in Class, and Behavior in Class are measured on a 1 to 5 Likert Scale. Statistics were generated from all available data.

3.2. Student Demographic and Test Score Information

Student demographic and achievement data came from district administrative records. Demographic data include gender, race/ethnicity, free- or reduced-price lunch (FRPL) eligibility, limited English proficiency (LEP) status, and special education (SPED) status. These records also included current- and prior-year test scores in math and English Language Arts (ELA) on state assessments, which we standardized within districts by grade, subject, and year using the entire sample of students.

The project also administered a low-stakes mathematics assessment to all students in the study. Internal consistency reliability is 0.82 or higher for each form across grade levels and school years ( Hickman, Fu, & Hill, 2012 ). We used this assessment in addition to high-stakes tests given that teacher effects on two outcomes that aim to capture similar underlying constructs (i.e., math achievement) provide a unique point of comparison when examining the relationship between teacher effects on student outcomes that are less closely related (i.e., math achievement versus attitudes and behaviors). Indeed, students’ high- and low-stake math test scores are correlated more strongly ( r = 0.70) than any other two outcomes (see Table 1 ). 5

3.3. Mathematics Lessons

Teachers’ mathematics lessons were captured over a three-year period, with an average of three lessons per teacher per year. 6 Trained raters scored these lessons on two established observational instruments, the CLASS and the MQI. Analyses of these same data show that items cluster into four main factors ( Blazar et al., 2015 ). The two dimensions from the CLASS instrument capture general teaching practices: Emotional Support focuses on teachers’ interactions with students and the emotional environment in the classroom, and is thought to increase students’ social and emotional development; and Classroom Organization focuses on behavior management and productivity of the lesson, and is thought to improve students’ self-regulatory behaviors ( Pianta & Hamre, 2009 ). 7 The two dimensions from the MQI capture mathematics-specific practices: Ambitious Mathematics Instruction focuses on the complexity of the tasks that teachers provide to their students and their interactions around the content, thus corresponding to the set of professional standards described by NCTM (1989 , 2014 ) and many elements contained within the Common Core State Standards for Mathematics ( National Governors Association Center for Best Practices, 2010 ); Mathematical Errors identifies any mathematical errors or imprecisions the teacher introduces into the lesson. Both dimensions from the MQI are linked to teachers’ mathematical knowledge for teaching and, in turn, to students’ math achievement ( Blazar, 2015 ; Hill et al., 2008 ; Hill, Schilling, & Ball, 2004 ). Correlations between dimensions range from roughly 0 (between Emotional Support and Mathematical Errors ) to 0.46 (between Emotional Support and Classroom Organization ; see Table 3 ).

Descriptive Statistics for CLASS and MQI Dimensions

Intraclass correlations were adjusted for the modal number of lessons. CLASS items (from Emotional Support and Classroom Organization) were scored on a scale from 1 to 7. MQI items (from Ambitious Instruction and Errors) were scored on a scale from 1 to 3. Statistics were generated from all available data.

We estimated reliability for these metrics by calculating the amount of variance in teacher scores that is attributable to the teacher (the intraclass correlation [ICC]), adjusted for the modal number of lessons. These estimates are: 0.53, 0.63, 0.74, and 0.56 for Emotional Support, Classroom Organization, Ambitious Mathematics Instruction , and Mathematical Errors , respectively (see Table 3 ). Though some of these estimates are lower than conventionally acceptable levels (0.7), they are consistent with those generated from similar studies ( Kane & Staiger, 2012 ). We standardized scores within the full sample of teachers to have a mean of zero and a standard deviation of one.

4. Empirical Strategy

4.1. estimating teacher effects on students’ attitudes and behaviors.

Like others who aim to examine the contribution of individual teachers to student outcomes, we began by specifying an education production function model of each outcome for student i in district d , school s , grade g , class c with teacher j at time t :

OUTCOME idsgict is used interchangeably for both math test scores and students’ attitudes and behaviors, which we modeled in separate equations as a cubic function of students’ prior achievement, A it −1 , in both math and ELA on the high-stakes district tests 8 ; demographic characteristics, X it , including gender, race, FRPL eligibility, SPED status, and LEP status; these same test-score variables and demographic characteristics averaged to the class level, X ¯ it c ; and district-by-grade-by-year fixed effects, τ dgt , that account for scaling of high-stakes test. The residual portion of the model can be decomposed into a teacher effect, µ j , which is our main parameter of interest and captures the contribution of teachers to student outcomes above and beyond factors already controlled for in the model; a class effect, δ jc , which is estimated by observing teachers over multiple school years; and a student-specific error term,. ε idsgjct 9

The key identifying assumption of this model is that teacher effect estimates are not biased by non-random sorting of students to teachers. Recent experimental ( Kane, McCaffrey, Miller, & Staiger, 2013 ) and quasi-experimental ( Chetty et al., 2014 ) analyses provide strong empirical support for this claim when student achievement is the outcome of interest. However, much less is known about bias and sorting mechanisms when other outcomes are used. For example, it is quite possible that students were sorted to teachers based on their classroom behavior in ways that were unrelated to their prior achievement. To address this possibility, we made two modifications to equation (1) . First, we included school fixed effects, ω s , to account for sorting of students and teachers across schools. This means that estimates rely only on between-school variation, which has been common practice in the literature estimating teacher effects on student achievement. In their review of this literature, Hanushek and Rivkin (2010) propose ignoring the between-school component because it is “surprisingly small” and because including this component leads to “potential sorting, testing, and other interpretative problems” (p. 268). Other recent studies estimating teacher effects on student outcomes beyond test scores have used this same approach ( Backes & Hansen, 2015 ; Gershenson, 2016 ; Jackson, 2012 ; Jennings & DiPrete, 2010 ; Ladd & Sorensen, 2015 ; Ruzek et al., 2015 ). Another important benefit of using school fixed effects is that this approach minimizes the possibility of reference bias in our self-reported measures ( West et al., 2016 ; Duckworth & Yeager, 2015 ). Differences in school-wide norms around behavior and effort may change the implicit standard of comparison (i.e. reference group) that students use to judge their own behavior and effort.

Restricting comparisons to other teachers and students within the same school minimizes this concern. As a second modification for models that predict each of our three student survey measures, we included OUTCOME it −1 on the right-hand side of the equation in addition to prior achievement – that is, when predicting students’ Behavior in Class , we controlled for students’ self-reported Behavior in Class in the prior year. 10 This strategy helps account for within-school sorting on factors other than prior achievement.

Using equation (1) , we estimated the variance of µ j , which is the stable component of teacher effects. We report the standard deviation of these estimates across outcomes. This parameter captures the magnitude of the variability of teacher effects. With the exception of teacher effects on students’ Happiness in Class , where survey items were not available in the first year of the study, we included δ jc in order to separate out the time-varying portion of teacher effects, combined with peer effects and any other class-level shocks. The fact that we are able to separate class effects from teacher effects is an important extension of prior studies examining teacher effects on outcomes beyond test scores, many of which only observed teachers at one point in time.

Following Chetty et al. (2011) , we estimated the magnitude of the variance of teacher effects using a direct, model-based estimate derived via restricted maximum likelihood estimation. This approach produces a consistent estimator for the true variance of teacher effects ( Raudenbush & Bryk, 2002 ). Calculating the variation across individual teacher effect estimates using Ordinary Least Squares regression would bias our variance estimates upward because it would conflate true variation with estimation error, particularly in instances where only a handful of students are attached to each teachers. Alternatively, estimating the variation in post-hoc predicted “shrunken” empirical Bayes estimates would bias our variance estimate downward relative to the size of the measurement error (Jacob & Lefgren, 2005).

4.2. Estimating Teaching Effects on Students’ Attitudes and Behaviors

We examined the contribution of teachers’ classroom practices to our set of student outcomes by estimating a variation of equation (1) :

This multi-level model includes the same set of control variables as above in order to account for the non-random sorting of students to teachers and for factors beyond teachers’ control that might influence each of our outcomes. We further included a vector of their teacher j ’s observation scores, OBSER VAT ^ ION l J , − t . The coefficients on these variables are our main parameters of interest and can be interpreted as the change in standard deviation units for each outcome associated with exposure to teaching practice one standard deviation above the mean.

One concern when relating observation scores to student survey outcomes is that they may capture the same behaviors. For example, teachers may receive credit on the Classroom Organization domain when their students demonstrate orderly behavior. In this case, we would have the same observed behaviors on both the left and right side of our equation relating instructional quality to student outcomes, which would inflate our teaching effect estimates. A related concern is that the specific students in the classroom may influence teachers’ instructional quality ( Hill et al., 2015 ; Steinberg & Garrett, 2016 ; Whitehurst, Chingos, & Lindquist, 2014 ). While the direction of bias is not as clear here – as either lesser- or higher-quality teachers could be sorted to harder to educate classrooms – this possibility also could lead to incorrect estimates. To avoid these sources of bias, we only included lessons captured in years other than those in which student outcomes were measured, denoted by – t in the subscript of OBSER VAT ^ ION l J , − t . To the extent that instructional quality varies across years, using out-of-year observation scores creates a lower-bound estimate of the true relationship between instructional quality and student outcomes. We consider this an important tradeoff to minimize potential bias. We used predicted shrunken observation score estimates that account for the fact that teachers contributed different numbers of lessons to the project, and fewer lessons could lead to measurement error in these scores ( Hill, Charalambous, & Kraft, 2012 ). 11

An additional concern for identification is the endogeneity of observed classroom quality. In other words, specific teaching practices are not randomly assigned to teachers. Our preferred analytic approach attempted to account for potential sources of bias by conditioning estimates of the relationship between one dimension of teaching practice and student outcomes on the three other dimensions. An important caveat here is that we only observed teachers’ instruction during math lessons and, thus, may not capture important pedagogical practices teachers used with these students when teaching other subjects. Including dimensions from the CLASS instrument, which are meant to capture instructional quality across subject areas ( Pianta & Hamre, 2009 ), helps account for some of this concern. However, given that we were not able to isolate one dimension of teaching quality from all others, we consider this approach as providing suggestive rather than conclusive evidence on the underlying causal relationship between teaching practice and students’ attitudes and behaviors.

4.3. Estimating the Relationship Between Teacher Effects Across Multiple Student Outcomes

In our third and final set of analyses, we examined whether teachers who are effective at raising math test scores are equally effective at developing students’ attitudes and behaviors. To do so, we drew on equation (1) to estimate µ̂ j for each outcome and teacher j . Following Chetty et al., 2014 ), we use post-hoc predicted “shrunken” empirical Bayes estimates of µ̂ j derived from equation (1) . Then, we generated a correlation matrix of these teacher effect estimates.

Despite attempts to increase the precision of these estimates through empirical Bayes estimation, estimates of individual teacher effects are measured with error that will attenuate these correlations ( Spearman, 1904 ). Thus, if we were to find weak to moderate correlations between different measures of teacher effectiveness, this could identify multidimensionality or could result from measurement challenges, including the reliability of individual constructs ( Chin & Goldhaber, 2015 ). For example, prior research suggests that different tests of students’ academic performance can lead to different teacher rankings, even when those tests measure similar underlying constructs ( Lockwood et al., 2007 ; Papay, 2011 ). To address this concern, we focus our discussion on relative rankings in correlations between teacher effect estimates rather than their absolute magnitudes. Specifically, we examine how correlations between teacher effects on two closely related outcomes (e.g., two math achievement tests) compare with correlations between teacher effects on outcomes that aim to capture different underlying constructs. In light of research highlighted above, we did not expect the correlation between teacher effects on the two math tests to be 1 (or, for that matter, close to 1). However, we hypothesized that these relationships should be stronger than the relationship between teacher effects on students’ math performance and effects on their attitudes and behaviors.

5.1. Do Teachers Impact Students’ Attitudes and Behaviors?

We begin by presenting results of the magnitude of teacher effects in Table 4 . Here, we observe sizable teacher effects on students’ attitudes and behaviors that are similar to teacher effects on students’ academic performance. Starting first with teacher effects on students’ academic performance, we find that a one standard deviation difference in teacher effectiveness is equivalent to a 0.17 sd or 0.18 sd difference in students’ math achievement. In other words, relative to an average teacher, teachers at the 84 th percentile of the distribution of effectiveness move the medium student up to roughly the 57 th percentile of math achievement. Notably, these findings are similar to those from other studies that also estimate within-school teacher effects in large administrative datasets ( Hanushek & Rivkin, 2010 ). This suggests that our use of school fixed effects with a more limited number of teachers observed within a given school does not appear to overly restrict our identifying variation. In Online Appendix A , where we present the magnitude of teacher effects from alternative model specifications, we show that results are robust to models that exclude school fixed effects or replace school fixed effects with observable school characteristics. Estimated teacher effects on students’ self-reported Self-Efficacy in Math and Behavior in Class are 0.14 sd and 0.15 sd, respectively. The largest teacher effects we observe are on students’ Happiness in Class , of 0.31 sd. Given that we do not have multiple years of data to separate out class effects for this measure, we interpret this estimate as the upward bound of true teacher effects on Happiness in Class. Rescaling this estimate by the ratio of teacher effects with and without class effects for Self-Efficacy in Math (0.14/0.19 = 0.74; see Online Appendix A ) produces an estimate of stable teacher effects on Happiness in Class of 0.23 sd, still larger than effects for other outcomes.

Teacher Effects on Students' Academic Performance, Attitudes, and Behaviors

Notes: Cells contain estimates from separate multi-level regression models.

All effects are statistically significant at the 0.05 level.

5.2. Do Specific Teaching Practices Impact Students’ Attitudes and Behaviors?

Next, we examine whether certain characteristics of teachers’ instructional practice help explain the sizable teacher effects described above. We present unconditional estimates in Table 5 Panel A, where the relationship between one dimension of teaching practice and student outcomes is estimated without controlling for the other three dimensions. Thus, cells contain estimates from separate regression models. In Panel B, we present conditional estimates, where all four dimensions of teaching quality are included in the same regression model. Here, columns contain estimates from separate regression models. We present all estimates as standardized effect sizes, which allows us to make comparisons across models and outcome measures. Unconditional and conditional estimates generally are quite similar. Therefore, we focus our discussion on our preferred conditional estimates.

Teaching Effects on Students' Academic Performance, Attitudes, and Behaviors

In Panel A, cells contain estimates from separate regression models. In Panel B, columns contain estimates from separate regression models, where estimates are conditioned on other teaching practices. All models control for student and class characteristics, school fixed effects, and district-by-grade-by-year fixed effects, and include and teacher random effects. Models predicting all outcomes except for Happiness in Class also include class random effects.

We find that students’ attitudes and behaviors are predicted by both general and content-specific teaching practices in ways that generally align with theory. For example, teachers’ Emotional Support is positively associated with the two closely related student constructs, Self-Efficacy in Math and Happiness in Class . Specifically, a one standard deviation increase in teachers’ Emotional Support is associated with a 0.14 sd increase in students’ Self-Efficacy in Math and a 0.37 sd increase in students’ Happiness in Class . These finding makes sense given that Emotional Support captures teacher behaviors such as their sensitivity to students, regard for students’ perspective, and the extent to which they create a positive climate in the classroom. As a point of comparison, these estimates are substantively larger than those between principal ratings of teachers’ ability to improve test scores and their actual ability to do so, which fall in the range of 0.02 sd and 0.08 sd ( Jacob & Lefgren, 2008 ; Rockoff, Staiger, Kane, & Taylor, 2012 ; Rockoff & Speroni, 2010 ).

We also find that Classroom Organization , which captures teachers’ behavior management skills and productivity in delivering content, is positively related to students’ reports of their own Behavior in Class (0.08 sd). This suggests that teachers who create an orderly classroom likely create a model for students’ own ability to self-regulate. Despite this positive relationship, we find that Classroom Organization is negatively associated with Happiness in Class (−0.23 sd), suggesting that classrooms that are overly focused on routines and management are negatively related to students’ enjoyment in class. At the same time, this is one instance where our estimate is sensitive to whether or not other teaching characteristics are included in the model. When we estimate the relationship between teachers’ Classroom Organization and students’ Happiness in Class without controlling for the three other dimensions of teaching quality, this estimate approaches 0 and is no longer statistically significant. 12 We return to a discussion of the potential tradeoffs between Classroom Organization and students’ Happiness in Class in our conclusion.

Finally, we find that the degree to which teachers commit Mathematical Errors is negatively related to students’ Self-Efficacy in Math (−0.09 sd) and Happiness in Class (−0.18 sd). These findings illuminate how a teacher’s ability to present mathematics with clarity and without serious mistakes is related to their students’ perceptions that they can complete math tasks and their enjoyment in class.

Comparatively, when predicting scores on both math tests, we only find one marginally significant relationship – between Mathematical Errors and the high-stakes math test (−0.02 sd). For two other dimensions of teaching quality, Emotional Support and Ambitious Mathematics Instruction , estimates are signed the way we would expect and with similar magnitudes, though they are not statistically significant. Given the consistency of estimates across the two math tests and our restricted sample size, it is possible that non-significant results are due to limited statistical power. 13 At the same time, even if true relationships exist between these teaching practices and students’ math test scores, they likely are weaker than those between teaching practices and students’ attitudes and behaviors. For example, we find that the 95% confidence intervals relating Classroom Emotional Support to Self-Efficacy in Math [0.068, 0.202] and Happiness in Class [0.162, 0.544] do not overlap with the 95% confidence intervals for any of the point estimates predicting math test scores. We interpret these results as indication that, still, very little is known about how specific classroom teaching practices are related to student achievement in math. 14

In Online Appendix B , we show that results are robust to a variety of different specifications, including (1) adjusting observation scores for characteristics of students in the classroom, (2) controlling for teacher background characteristics (i.e., teaching experience, math content knowledge, certification pathway, education), and (3) using raw out-of-year observation scores (rather than shrunken scores). This suggests that our approach likely accounts for many potential sources of bias in our teaching effect estimates.

5.3. Are Teachers Equally Effective at Raising Different Student Outcomes?

In Table 6 , we present correlations between teacher effects on each of our student outcomes. The fact that teacher effects are measured with error makes it difficult to estimate the precise magnitude of these correlations. Instead, we describe relative differences in correlations, focusing on the extent to which teacher effects within outcome type – i.e., teacher effects on the two math achievement tests or effects on students’ attitudes and behaviors – are similar or different from correlations between teacher effects across outcome type. We illustrate these differences in Figure 1 , where Panel A presents scatter plots of these relationships between teacher effects within outcome type and Panel B does the same across outcome type. Recognizing that not all of our survey outcomes are meant to capture the same underlying construct, we also describe relative differences in correlations between teacher effects on these different measures. In Online Appendix C , we find that an extremely conservative adjustment that scales correlations by the inverse of the square root of the product of the reliabilities leads to a similar overall pattern of results.

Scatter plots of teacher effects across outcomes. Solid lines represent the best-fit regression line.

Correlations Between Teacher Effects on Students' Academic Performance, Attitudes, and Behaviors

Standard errors in parentheses. See Table 4 for sample sizes used to calculate teacher effect estimates. The sample for each correlation is the minimum number of teachers between the two measures.

Examining the correlations of teacher effect estimates reveals that individual teachers vary considerably in their ability to impact different student outcomes. As hypothesized, we find the strongest correlations between teacher effects within outcome type. Similar to Corcoran, Jennings, and Beveridge (2012) , we estimate a correlation of 0.64 between teacher effects on our high- and low-stakes math achievement tests. We also observe a strong correlation of 0.49 between teacher effects on two of the student survey measures, students’ Behavior in Class and Self-Efficacy in Math . Comparatively, the correlations between teacher effects across outcome type are much weaker. Examining the scatter plots in Figure 1 , we observe much more dispersion around the best-fit line in Panel B than in Panel A. The strongest relationship we observe across outcome types is between teacher effects on the low-stakes math test and effects on Self-Efficacy in Math ( r = 0.19). The lower bound of the 95% confidence interval around the correlation between teacher effects on the two achievement measures [0.56, 0.72] does not overlap with the 95% confidence interval of the correlation between teacher effects on the low-stakes math test and effects on Self-Efficacy in Math [−0.01, 0.39], indicating that these two correlations are substantively and statistically significantly different from each other. Using this same approach, we also can distinguish the correlation describing the relationship between teacher effects on the two math tests from all other correlations relating teacher effects on test scores to effects on students’ attitudes and behaviors. We caution against placing too much emphasis on the negative correlations between teacher effects on test scores and effects on Happiness in Class ( r = −0.09 and −0.21 for the high- and low-stakes tests, respectively). Given limited precision of this relationship, we cannot reject the null hypothesis of no relationship or rule out weak, positive or negative correlations among these measures.

Although it is useful to make comparisons between the strength of the relationships between teacher effects on different measures of students’ attitudes and behaviors, measurement error limits our ability to do so precisely. At face value, we find correlations between teacher effects on Happiness in Class and effects on the two other survey measures ( r = 0.26 for Self-Efficacy in Math and 0.21 for Behavior in Class ) that are weaker than the correlation between teacher effects on Self-Efficacy in Math and effects on Behavior in Class described above ( r = 0.49). One possible interpretation of these findings is that teachers who improve students’ Happiness in Class are not equally effective at raising other attitudes and behaviors. For example, teachers might make students happy in class in unconstructive ways that do not also benefit their self-efficacy or behavior. At the same time, these correlations between teacher effects on Happiness in Class and the other two survey measures have large confidence intervals, likely due to imprecision in our estimate of teacher effects on Happiness in Class . Thus, we are not able to distinguish either correlation from the correlation between teacher effects on Behavior in Class and effects on Self-Efficacy in Math .

6. Discussion and Conclusion

6.1. relationship between our findings and prior research.

The teacher effectiveness literature has profoundly shaped education policy over the last decade and has served as the catalyst for sweeping reforms around teacher recruitment, evaluation, development, and retention. However, by and large, this literature has focused on teachers’ contribution to students’ test scores. Even research studies such as the Measures of Effective Teaching project and new teacher evaluation systems that focus on “multiple measures” of teacher effectiveness ( Center on Great Teachers and Leaders, 2013 ; Kane et al., 2013 ) generally attempt to validate other measures, such as observations of teaching practice, by examining their relationship to estimates of teacher effects on students’ academic performance.

Our study extends an emerging body of research examining the effect of teachers on student outcomes beyond test scores. In many ways, our findings align with conclusions drawn from previous studies that also identify teacher effects on students’ attitudes and behaviors ( Jennings & DiPrete, 2010 ; Kraft & Grace, 2016 ; Ruzek et al., 2015 ), as well as weak relationships between different measures of teacher effectiveness ( Gershenson, 2016 ; Jackson, 2012 ; Kane & Staiger, 2012 ). To our knowledge, this study is the first to identify teacher effects on measures of students’ self-efficacy in math and happiness in class, as well as on a self-reported measure of student behavior. These findings suggest that teachers can and do help develop attitudes and behaviors among their students that are important for success in life. By interpreting teacher effects alongside teaching effects, we also provide strong face and construct validity for our teacher effect estimates. We find that improvements in upper-elementary students’ attitudes and behaviors are predicted by general teaching practices in ways that align with hypotheses laid out by instrument developers ( Pianta & Hamre, 2009 ). Findings linking errors in teachers’ presentation of math content to students’ self-efficacy in math, in addition to their math performance, also are consistent with theory ( Bandura et al., 1996 ). Finally, the broad data collection effort from NCTE allows us to examine relative differences in relationships between measures of teacher effectiveness, thus avoiding some concerns about how best to interpret correlations that differ substantively across studies ( Chin & Goldhaber, 2015 ). We find that correlations between teacher effects on student outcomes that aim to capture different underlying constructs (e.g., math test scores and behavior in class) are weaker than correlations between teacher effects on two outcomes that are much more closely related (e.g., math achievement).

6.2. Implications for Policy

These findings can inform policy in several key ways. First, our findings may contribute to the recent push to incorporate measures of students’ attitudes and behaviors – and teachers’ ability to improve these outcomes – into accountability policy (see Duckworth, 2016 ; Miller, 2015 ; Zernike, 2016 for discussion of these efforts in the press). After passage of the Every Student Succeeds Act (ESSA), states now are required to select a nonacademic indicator with which to assess students’ success in school ( ESSA, 2015 ). Including measures of students’ attitudes and behaviors in accountability or evaluation systems, even with very small associated weights, could serve as a strong signal that schools and educators should value and attend to developing these skills in the classroom.

At the same time, like other researchers ( Duckworth & Yeager, 2015 ), we caution against a rush to incorporate these measures into high-stakes decisions. The science of measuring students’ attitudes and behaviors is relatively new compared to the long history of developing valid and reliable assessments of cognitive aptitude and content knowledge. Most existing measures, including those used in this study, were developed for research purposes rather than large-scale testing with repeated administrations. Open questions remain about whether reference bias substantially distorts comparisons across schools. Similar to previous studies, we include school fixed effects in all of our models, which helps reduce this and other potential sources of bias. However, as a result, our estimates are restricted to within-school comparisons of teachers and cannot be applied to inform the type of across-school comparisons that districts typically seek to make. There also are outstanding questions regarding the susceptibility of these measures to “survey” coaching when high-stakes incentives are attached. Such incentives likely would render teacher or self-assessments of students’ attitudes and behaviors inappropriate. Some researchers have started to explore other ways to capture students’ attitudes and behaviors, including objective performance-based tasks and administrative proxies such as attendance, suspensions, and participation in extracurricular activities ( Hitt, Trivitt, & Cheng, 2016 ; Jackson, 2012 ; Whitehurst, 2016 ). This line of research shows promise but still is in its early phases. Further, although our modeling strategy aims to reduce bias due to non-random sorting of students to teachers, additional evidence is needed to assess the validity of this approach. Without first addressing these concerns, we believe that adding untested measures into accountability systems could lead to superficial and, ultimately, counterproductive efforts to support the positive development of students’ attitudes and behaviors.

An alternative approach to incorporating teacher effects on students’ attitudes and behaviors into teacher evaluation may be through observations of teaching practice. Our findings suggest that specific domains captured on classroom observation instruments (i.e., Emotional Support and Classroom Organization from the CLASS and Mathematical Errors from the MQI) may serve as indirect measures of the degree to which teachers impact students’ attitudes and behaviors. One benefit of this approach is that districts commonly collect related measures as part of teacher evaluation systems ( Center on Great Teachers and Leaders, 2013 ), and such measures are not restricted to teachers who work in tested grades and subjects.

Similar to Whitehurst (2016) , we also see alternative uses of teacher effects on students’ attitudes and behaviors that fall within and would enhance existing school practices. In particular, measures of teachers’ effectiveness at improving students’ attitudes and behaviors could be used to identify areas for professional growth and connect teachers with targeted professional development. This suggestion is not new and, in fact, builds on the vision and purpose of teacher evaluation described by many other researchers ( Darling-Hammond, 2013 ; Hill & Grossman, 2013 ; Papay, 2012 ). However, in order to leverage these measures for instructional improvement, we add an important caveat: performance evaluations – whether formative or summative – should avoid placing teachers into a single performance category whenever possible. Although many researchers and policymakers argue for creating a single weighted composite of different measures of teachers’ effectiveness ( Center on Great Teachers and Leaders, 2013 ; Kane et al., 2013 ), doing so likely oversimplifies the complex nature of teaching. For example, a teacher who excels at developing students’ math content knowledge but struggles to promote joy in learning or students’ own self-efficacy in math is a very different teacher than one who is middling across all three measures. Looking at these two teachers’ composite scores would suggest they are similarly effective. A single overall evaluation score lends itself to a systematized process for making binary decisions such as whether to grant teachers tenure, but such decisions would be better informed by recognizing and considering the full complexity of classroom practice.

We also see opportunities to maximize students’ exposure to the range of teaching skills we examine through strategic teacher assignments. Creating a teacher workforce skilled in most or all areas of teaching practice is, in our view, the ultimate goal. However, this goal likely will require substantial changes to teacher preparation programs and curriculum materials, as well as new policies around teacher recruitment, evaluation, and development. In middle and high schools, content-area specialization or departmentalization often is used to ensure that students have access to teachers with skills in distinct content areas. Some, including the National Association of Elementary School Principals, also see this as a viable strategy at the elementary level ( Chan & Jarman, 2004 ). Similar approaches may be taken to expose students to a collection of teachers who together can develop a range of academic skills, attitudes and behaviors. For example, when configuring grade-level teams, principals may pair a math teacher who excels in her ability to improve students’ behavior with an ELA or reading teacher who excels in his ability to improve students’ happiness and engagement. Viewing teachers as complements to each other may help maximize outcomes within existing resource constraints.

Finally, we consider the implications of our findings for the teaching profession more broadly. While our findings lend empirical support to research on the multidimensional nature of teaching ( Cohen, 2011 ; Lampert, 2001 ; Pianta & Hamre, 2009 ), we also identify tensions inherent in this sort of complexity and potential tradeoffs between some teaching practices. In our primary analyses, we find that high-quality instruction around classroom organization is positively related to students’ self-reported behavior in class but negatively related to their happiness in class. Our results here are not conclusive, as the negative relationship between classroom organization and students’ happiness in class is sensitive to model specification. However, if there indeed is a negative causal relationship, it raises questions about the relative benefits of fostering orderly classroom environments for learning versus supporting student engagement by promoting positive experiences with schooling. Our own experience as educators and researchers suggests this need not be a fixed tradeoff. Future research should examine ways in which teachers can develop classroom environments that engender both constructive classroom behavior and students’ happiness in class. As our study draws on a small sample of students who had current and prior-year scores for Happiness in Class , we also encourage new studies with greater statistical power that may be able to uncover additional complexities (e.g., non-linear relationships) in these sorts of data.

Our findings also demonstrate a need to integrate general and more content-specific perspectives on teaching, a historical challenge in both research and practice ( Grossman & McDonald, 2008 ; Hamre et al., 2013 ). We find that both math-specific and general teaching practices predict a range of student outcomes. Yet, particularly at the elementary level, teachers’ math training often is overlooked. Prospective elementary teachers often gain licensure without taking college-level math classes; in many states, they do not need to pass the math sub-section of their licensure exam in order to earn a passing grade overall ( Epstein & Miller, 2011 ). Striking the right balance between general and content-specific teaching practices is not a trivial task, but it likely is a necessary one.

For decades, efforts to improve the quality of the teacher workforce have focused on teachers’ abilities to raise students’ academic achievement. Our work further illustrates the potential and importance of expanding this focus to include teachers’ abilities to promote students’ attitudes and behaviors that are equally important for students’ long-term success.

Supplementary Material

Acknowledgments.

The research reported here was supported in part by the Institute of Education Sciences, U.S. Department of Education, through Grant R305C090023 to the President and Fellows of Harvard College to support the National Center for Teacher Effectiveness. The opinions expressed are those of the authors and do not represent views of the Institute or the U.S. Department of Education. Additional support came from the William T. Grant Foundation, the Albert Shanker Institute, and Mathematica Policy Research’s summer fellowship.

Appendix Table 1

Factor Loadings for Items from the Student Survey

Notes: Estimates drawn from all available data. Loadings of roughly 0.4 or higher are highlighted to identify patterns.

1 Although student outcomes beyond test scores often are referred to as “non-cognitive” skills, our preference, like others ( Duckworth & Yeager, 2015 ; Farrington et al., 2012 ), is to refer to each competency by name. For brevity, we refer to them as “attitudes and behaviors,” which closely characterizes the measures we focus on in this paper.

2 Analyses below include additional subsamples of teachers and students. In analyses that predict students’ survey response, we included between 51 and 111 teachers and between 548 and 1,529 students. This range is due to the fact that some survey items were not available in the first year of the study. Further, in analyses relating domains of teaching practice to student outcomes, we further restricted our sample to teachers who themselves were part of the study for more than one year, which allowed us to use out-of-year observation scores that were not confounded with the specific set of students in the classroom. This reduced our analysis samples to between 47 and 93 teachers and between 517 and 1,362 students when predicting students’ attitudes and behaviors, and 196 teachers and 8,660 students when predicting math test scores. Descriptive statistics and formal comparisons of other samples show similar patterns as those presented in Table 1 .

3 We conducted factor analyses separately by year, given that additional items were added in the second and third years to help increase reliability. In the second and third years, each of the two factors has an eigenvalue above one, a conventionally used threshold for selecting factors ( Kline, 1994 ). Even though the second factor consists of three items that also have loadings on the first factor between 0.35 and 0.48 – often taken as the minimum acceptable factor loading ( Field, 2013 ; Kline, 1994 ) – this second factor explains roughly 20% more of the variation across teachers and, therefore, has strong support for a substantively separate construct ( Field, 2013 ; Tabachnick & Fidell, 2001 ). In the first year of the study, the eigenvalue on this second factor is less strong (0.78), and the two items that load onto it also load onto the first factor.

4 Depending on the outcome, between 4% and 8% of students were missing a subset of items from survey scales. In these instances, we created final scores by averaging across all available information.

5 Coding of items from both the low- and high-stakes tests also identify a large degree of overlap in terms of content coverage and cognitive demand ( Lynch, Chin, & Blazar, 2015 ). All tests focused most on numbers and operations (40% to 60%), followed by geometry (roughly 15%), and algebra (15% to 20%). By asking students to provide explanations of their thinking and to solve non-routine problems such as identifying patterns, the low-stakes test also was similar to the high-stakes tests in two districts; in the other two districts, items often asked students to execute basic procedures.

6 As described by Blazar (2015) , capture occurred with a three-camera, digital recording device and lasted between 45 and 60 minutes. Teachers were allowed to choose the dates for capture in advance and directed to select typical lessons and exclude days on which students were taking a test. Although it is possible that these lessons were unique from a teachers’ general instruction, teachers did not have any incentive to select lessons strategically as no rewards or sanctions were involved with data collection or analyses. In addition, analyses from the MET project indicate that teachers are ranked almost identically when they choose lessons themselves compared to when lessons are chosen for them ( Ho & Kane, 2013 ).

7 Developers of the CLASS instrument identify a third dimension, Classroom Instructional Support . Factor analyses of data used in this study showed that items from this dimension formed a single construct with items from Emotional Support ( Blazar et al., 2015 ). Given theoretical overlap between Classroom Instructional Support and dimensions from the MQI instrument, we excluded these items from our work and focused only on Classroom Emotional Support.

8 We controlled for prior-year scores only on the high-stakes assessments and not on the low-stakes assessment for three reasons. First, including prior low-stakes test scores would reduce our full sample by more than 2,200 students. This is because the assessment was not given to students in District 4 in the first year of the study (N = 1,826 students). Further, an additional 413 students were missing fall test scores given that they were not present in class on the day it was administered. Second, prior-year scores on the high- and low-stakes test are correlated at 0.71, suggesting that including both would not help to explain substantively more variation in our outcomes. Third, sorting of students to teachers is most likely to occur based on student performance on the high-stakes assessments since it was readily observable to schools; achievement on the low-stakes test was not.

9 An alternative approach would be to specify teacher effects as fixed, rather than random, which relaxes the assumption that teacher assignment is uncorrelated with factors that also predict student outcomes ( Guarino, Maxfield, Reckase, Thompson, & Wooldridge, 2015 ). Ultimately, we prefer the random effects specification for three reasons. First, it allows us to separate out teacher effects from class effects by including a random effect for both in our model. Second, this approach allows us to control for a variety of variables that are dropped from the model when teacher fixed effects also are included. Given that all teachers in our sample remained in the same school from one year to the next, school fixed effects are collinear with teacher fixed effects. In instances where teachers had data for only one year, class characteristics and district-by-grade-by-year fixed effects also are collinear with teacher fixed effects. Finally, and most importantly, we find that fixed and random effects specifications that condition on students’ prior achievement and demographic characteristics return almost identical teacher effect estimates. When comparing teacher fixed effects to the “shrunken” empirical Bayes estimates that we employ throughout the paper, we find correlations between 0.79 and 0.99. As expected, the variance of the teacher fixed effects is larger than the variance of teacher random effects, differing by the shrinkage factor. When we instead calculate teacher random effects without shrinkage by averaging student residuals to the teacher level (i.e., “teacher average residuals”; see Guarino et al, 2015 for a discussion of this approach) they are almost identical to the teacher fixed effects estimates. Correlations are 0.99 or above across outcome measures, and unstandardized regression coefficients that retain the original scale of each measure range from 0.91 sd to 0.99 sd.

10 Adding prior survey responses to the education production function is not entirely analogous to doing so with prior achievement. While achievement outcomes have roughly the same reference group across administrations, the surveys do not. This is because survey items often asked about students’ experiences “in this class.” All three Behavior in Class items and all five Happiness in Class items included this or similar language, as did five of the 10 items from Self-Efficacy in Math . That said, moderate year-to-year correlations of 0.39, 0.38, and 0.53 for Self-Efficacy in Math , Happiness in Class , and Behavior in Class , respectively, suggest that these items do serve as important controls. Comparatively, year-to-year correlations for the high- and low-stakes tests are 0.75 and 0.77.

11 To estimate these scores, we specified the following hierarchical linear model separately for each school year: OBSER VAT ^ ION lj , − t = γ j + ε ljt The outcome is the observation score for lesson l from teacher j in years other than t ; γ j is a random effect for each teacher, and ε ljt is the residual. For each domain of teaching practice and school year, we utilized standardized estimates of the teacher-level residual as each teacher’s observation score in that year. Thus, scores vary across time. In the main text, we refer to these teacher-level residual as OBSER VAT ^ ION l J , − t rather than γ ̂ J for ease of interpretation for readers.

12 One explanation for these findings is that the relationship between teachers’ Classroom Organization and students’ Happiness in Class is non-liner. For example, it is possible that students’ happiness increases as the class becomes more organized, but then begins to decrease in classrooms with an intensive focus on order and discipline. To explore this possibility, we first examined the scatterplot of the relationship between teachers’ Classroom Organization and teachers’ ability to improve students’ Happiness in Class . Next, we re-estimated equation (2) including a quadratic, cubic, and quartic specification of teachers’ Classroom Organization scores. In both sets of analyses, we found no evidence for a non-linear relationship. Given our small sample size and limited statistical power, though, we suggest that this may be a focus of future research.

13 In similar analyses in a subset of the NCTE data, Blazar (2015) did find a statistically significant relationship between Ambitious Mathematics Instruction and the low-stakes math test of 0.11 sd. The 95% confidence interval around that point estimate overlaps with the 95% confidence interval relating Ambitious Mathematics Instruction to the low-stakes math test in this analysis. Estimates of the relationship between the other three domains of teaching practice and low-stakes math test scores were of smaller magnitude and not statistically significant. Differences between the two studies likely emerge from the fact that we drew on a larger sample with an additional district and year of data, as well as slight modifications to our identification strategy.

14 When we adjusted p -values for estimates presented in Table 5 to account for multiple hypothesis testing using both the Šidák and Bonferroni algorithms ( Dunn, 1961 ; Šidák, 1967 ), relationships between Emotional Support and both Self-Efficacy in Math and Happiness in Class , as well as between Mathematical Errors and Self-Efficacy in Math remained statistically significant.

Contributor Information

David Blazar, Harvard Graduate School of Education.

Matthew A. Kraft, Brown University.

• Achenbach TM, McConaughy SH, Howell CT. Child/adolescent behavioral and emotional problems: implications of cross-informant correlations for situational specificity. Psychological Bulletin. 1987; 101 (2):213. [ PubMed ] [ Google Scholar ]
• Backes B, Hansen M. Working Paper 146. Washington, D C: National Center for Analysis of Longitudinal in Education Research; 2015. Teach for America impact estimates on nontested student outcomes. Retrieved from http://www.caldercenter.org/sites/default/files/WP&%20146.pdf . [ Google Scholar ]
• Bandura A, Barbaranelli C, Caprara GV, Pastorelli C. Multifaceted impact of self-efficacy beliefs on academic functioning. Child Development. 1996:1206–1222. [ PubMed ] [ Google Scholar ]
• Baron J. Personality and intelligence. In: Sternberg RJ, editor. Handbook of human intelligence. New York: Cambridge University Press; 1982. pp. 308–351. [ Google Scholar ]
• Blazar D. Effective teaching in elementary mathematics: Identifying classroom practices that support student achievement. Economics of Education Review. 2015; 48 :16–29. [ Google Scholar ]
• Blazar D, Braslow D, Charalambous CY, Hill HC. Working Paper. Cambridge, MA: National Center for Teacher Effectiveness; 2015. Attending to general and content-specific dimensions of teaching: Exploring factors across two observation instruments. Retrieved from http://scholar.harvard.edu/files/david_blazar/files/blazar_et_al_attending_to_general_and_content_specific_dimensions_of_teaching.pdf . [ Google Scholar ]
• Borghans L, Duckworth AL, Heckman JJ, Ter Weel B. The economics and psychology of personality traits. Journal of Human Resources. 2008; 43 (4):972–1059. [ Google Scholar ]
• Burchinal M, Howes C, Pianta R, Bryant D, Early D, Clifford R, Barbarin O. Predicting child outcomes at the end of kindergarten from the quality of pre-kindergarten teacher-child interactions and instruction. Applied Developmental Science. 2008; 12 (3):140–153. [ Google Scholar ]
• Center on Great Teachers and Leaders. Databases on state teacher and principal policies. 2013 Retrieved from: http://resource.tqsource.org/stateevaldb .
• Chan TC, Jarman D. Departmentalize elementary schools. Principal. 2004; 84 (1):70–72. [ Google Scholar ]
• Chetty R, Friedman JN, Hilger N, Saez E, Schanzenbach D, Yagan D. How does your kindergarten classroom affect your earnings? Evidence from Project STAR. Quarterly Journal of Economics. 2011; 126 (4):1593–1660. [ PubMed ] [ Google Scholar ]
• Chetty R, Friedman JN, Rockoff JE. Measuring the impacts of teachers I: Evaluating Bias in Teacher Value-Added Estimates. American Economic Review. 2014; 104 (9):2593–2632. [ Google Scholar ]
• Chin M, Goldhaber D. Working Paper. Cambridge, MA: National Center for Teacher Effectiveness; 2015. Exploring explanations for the “weak” relationship between value added and observation-based measures of teacher performance. Retrieved from: http://cepr.harvard.edu/files/cepr/files/sree2015_simulation_working_paper.pdf?m=1436541369 . [ Google Scholar ]
• Cohen DK. Teaching and its predicaments. Cambridge, MA: Harvard University Press; 2011. [ Google Scholar ]
• Corcoran SP, Jennings JL, Beveridge AA. Teacher effectiveness on high- and low-stakes tests. 2012 Unpublished manuscript. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.269.5537&rep=rep1&type=pdf .
• Council of the Great City Schools. Beating the odds: Analysis of student performance on state assessments results from the 2012–2013 school year. Washington, DC: Author; 2013. [ Google Scholar ]
• Darling-Hammond L. Getting teacher evaluation right: What really matters for effectiveness and improvement. New York: Teachers College Press; 2013. [ Google Scholar ]
• Diener E. Subjective well-being: The science of happiness and a proposal for a national index. American Psychologist. 2000; 55 (1):34–43. [ PubMed ] [ Google Scholar ]
• Downer JT, Rimm-Kaufman S, Pianta RC. How do classroom conditions and children's risk for school problems contribute to children's behavioral engagement in learning? School Psychology Review. 2007; 36 (3):413–432. [ Google Scholar ]
• Duckworth A. Don’t grade schools on grit. The New York Times. 2016 Mar 26; Retrieved from http://www.nytimes.com/2016/03/27/opinion/sunday/dont-grade-schools-on-grit.html .
• Duckworth AL, Peterson C, Matthews MD, Kelly DR. Grit: Perseverance and passion for long-term goals. Journal of Personality and Social Psychology. 2007; 92 (6):1087–1101. [ PubMed ] [ Google Scholar ]
• Duckworth AL, Quinn PD, Tsukayama E. What No Child Left Behind leaves behind: The roles of IQ and self-control in predicting standardized achievement test scores and report card grades. Journal of Educational Psychology. 2012; 104 (2):439–451. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Duckworth AL, Yeager DS. Measurement matters: Assessing personal qualities other than cognitive ability for educational purposes. Educational Researcher. 2015; 44 (4):237–251. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Dunn OJ. Multiple comparisons among means. Journal of the American Statistical Association. 1961; 56 (293):52–64. [ Google Scholar ]
• Epstein D, Miller RT. Slow off the mark: Elementary school teachers and the crisis in science, technology, engineering, and math education. Washington, DC: Center for American Progress; 2011. [ Google Scholar ]
• The Every Student Succeeds Act. Public Law 114-95, 114th Cong., 1st sess. 2015 Dec 10; available at https://www.congress.gov/bill/114th-congress/senate-bill/1177/text .
• Farrington CA, Roderick M, Allensworth E, Nagaoka J, Keyes TS, Johnson DW, Beechum NO. Teaching adolescents to become learners: The role of non-cognitive factors in shaping school performance, a critical literature review. Chicago: University of Chicago Consortium on Chicago School Reform; 2012. [ Google Scholar ]
• Field A. Discovering statistics using IBM SPSS statistics. 4. London: SAGE publications; 2013. [ Google Scholar ]
• Gershenson S. Linking teacher quality, student attendance, and student achievement. Education Finance and Policy. 2016; 11 (2):125–149. [ Google Scholar ]
• Goodman R. Psychometric properties of the strengths and difficulties questionnaire. Journal of the American Academy of Child & Adolescent Psychiatry. 2001; 40 (11):1337–1345. [ PubMed ] [ Google Scholar ]
• Grossman P, McDonald M. Back to the future: Directions for research in teaching and teacher education. American Educational Research Journal. 2008; 45 :184–205. [ Google Scholar ]
• Guarino CM, Maxfield M, Reckase MD, Thompson PN, Wooldridge JM. An evaluation of Empirical Bayes’ estimation of value-added teacher performance measures. Journal of Educational and Behavioral Statistics. 2015; 40 (2):190–222. [ Google Scholar ]
• Hafen CA, Hamre BK, Allen JP, Bell CA, Gitomer DH, Pianta RC. Teaching through interactions in secondary school classrooms: Revisiting the factor structure and practical application of the classroom assessment scoring system–secondary. The Journal of Early Adolescence. 2015; 35 (5–6):651–680. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hamre B, Hatfield B, Pianta R, Jamil F. Evidence for general and domain-specific elements of teacher–child interactions: Associations with preschool children's development. Child Development. 2014; 85 (3):1257–1274. [ PubMed ] [ Google Scholar ]
• Hamre BK, Pianta RC. Early teacher–child relationships and the trajectory of children's school outcomes through eighth grade. Child Development. 2001; 72 (2):625–638. [ PubMed ] [ Google Scholar ]
• Hamre BK, Pianta RC, Downer JT, DeCoster J, Mashburn AJ, Jones SM, Brackett MA. Teaching through interactions: Testing a developmental framework of teacher effectiveness in over 4,000 classrooms. The Elementary School Journal. 2013; 113 (4):461–487. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hanushek EA, Rivkin SG. Generalizations about using value-added measures of teacher quality. American Economic Review. 2010; 100 (2):267–271. [ Google Scholar ]
• Hickman JJ, Fu J, Hill HC. Technical report: Creation and dissemination of upper-elementary mathematics assessment modules. Princeton, NJ: Educational Testing Service; 2012. [ Google Scholar ]
• Hill HC, Blazar D, Lynch K. Resources for teaching: Examining personal and institutional predictors of high-quality instruction. AERA Open. 2015; 1 (4):1–23. [ Google Scholar ]
• Hill HC, Blunk ML, Charalambous CY, Lewis JM, Phelps GC, Sleep L, Ball DL. Mathematical knowledge for teaching and the mathematical quality of instruction: An exploratory study. Cognition and Instruction. 2008; 26 (4):430–511. [ Google Scholar ]
• Hill HC, Charalambous CY, Kraft MA. When rater reliability is not enough teacher observation systems and a case for the generalizability study. Educational Researcher. 2012; 41 (2):56–64. [ Google Scholar ]
• Hill HC, Grossman P. Learning from teacher observations: Challenges and opportunities posed by new teacher evaluation systems. Harvard Educational Review. 2013; 83 (2):371–384. [ Google Scholar ]
• Hill HC, Schilling SG, Ball DL. Developing measures of teachers’ mathematics knowledge for teaching. Elementary School Journal. 2004; 105 :11–30. [ Google Scholar ]
• Hitt C, Trivitt J, Cheng A. When you say nothing at all: The predictive power of student effort on surveys. Economics of Education Review. 2016; 52 :105–119. [ Google Scholar ]
• Ho AD, Kane TJ. The reliability of classroom observations by school personnel. Seattle, WA: Measures of Effective Teaching Project, Bill and Melinda Gates Foundation; 2013. [ Google Scholar ]
• Jackson CK. NBER Working Paper No. 18624. Cambridge, MA: National Bureau for Economic Research; 2012. Non-cognitive ability, test scores, and teacher quality: Evidence from ninth grade teachers in North Carolina. [ Google Scholar ]
• Jacob BA, Lefgren L. Can principals identify effective teachers? Evidence on subjective performance evaluation in education. Journal of Labor Economics. 2008; 20 (1):101–136. [ Google Scholar ]
• Jennings JL, DiPrete TA. Teacher effects on social and behavioral skills in early elementary school. Sociology of Education. 2010; 83 (2):135–159. [ Google Scholar ]
• John OP, Srivastava S. The Big Five trait taxonomy: History, measurement, and theoretical perspectives. Handbook of personality: Theory and research. 1999; 2 (1999):102–138. [ Google Scholar ]
• Kane TJ, McCaffrey DF, Miller T, Staiger DO. Have we identified effective teachers? Validating measures of effective teaching using random assignment. Seattle, WA: Measures of Effective Teaching Project, Bill and Melinda Gates Foundation; 2013. [ Google Scholar ]
• Kane TJ, Staiger DO. Gathering feedback for teaching. Seattle, WA: Measures of Effective Teaching Project, Bill and Melinda Gates Foundation; 2012. [ Google Scholar ]
• King RB, McInerney DM, Ganotice FA, Villarosa JB. Positive affect catalyzes academic engagement: Cross-sectional, longitudinal, and experimental evidence. Learning and Individual Differences. 2015; 39 :64–72. [ Google Scholar ]
• Kline P. An easy guide to factor analysis. London: Routledge; 1994. [ Google Scholar ]
• Kraft MA, Grace S. Working Paper. Providence, RI: Brown University; 2016. Teaching for tomorrow’s economy? Teacher effects on complex cognitive skills and social-emotional competencies. Retrieved from http://scholar.harvard.edu/files/mkraft/files/teaching_for_tomorrows_economy_-_final_public.pdf . [ Google Scholar ]
• Koedel C. Teacher quality and dropout outcomes in a large, urban school district. Journal of Urban Economics. 2008; 64 (3):560–572. [ Google Scholar ]
• Ladd HF, Sorensen LC. Working Paper No. 112. Washington, D C: National Center for Analysis of Longitudinal in Education Research; 2015. Returns to teacher experience: Student achievement and motivation in middle school. Retrieved from http://www.caldercenter.org/sites/default/files/WP%20112%20Update_0.pdf . [ Google Scholar ]
• Lampert M. Teaching problems and the problems of teaching. Yale University Press; 2001. [ Google Scholar ]
• Lockwood JR, McCaffrey DF, Hamilton LS, Stecher B, Le V, Martinez JF. The sensitivity of value-added teacher effect estimates to different mathematics achievement measures. Journal of Educational Measurement. 2007; 44 (1):47–67. [ Google Scholar ]
• Luckner AE, Pianta RC. Teacher-student interactions in fifth grade classrooms: Relations with children's peer behavior. Journal of Applied Developmental Psychology. 2011; 32 (5):257–266. [ Google Scholar ]
• Lynch K, Chin M, Blazar D. Working Paper. Cambridge, MA: National Center for Teacher Effectiveness; 2015. Relationship between observations of elementary teacher mathematics instruction and student achievement: Exploring variability across districts. [ Google Scholar ]
• Lyubomirsky S, King L, Diener E. The benefits of frequent positive affect: Does happiness lead to success? Psychological Bulletin. 2005; 131 (6):803–855. [ PubMed ] [ Google Scholar ]
• Mashburn AJ, Pianta RC, Hamre BK, Downer JT, Barbarin OA, Bryant D, Howes C. Measures of classroom quality in prekindergarten and children's development of academic, language, and social skills. Child Development. 2008; 79 (3):732–749. [ PubMed ] [ Google Scholar ]
• Mihaly K, McCaffrey DF, Staiger DO, Lockwood JR. A composite estimator of effective teaching. Seattle, WA: Measures of Effective Teaching Project, Bill and Melinda Gates Foundation; 2013. [ Google Scholar ]
• Miles SB, Stipek D. Contemporaneous and longitudinal associations between social behavior and literacy achievement in a sample of low-income elementary school children. Child Development. 2006; 77 (1):103–117. [ PubMed ] [ Google Scholar ]
• Miller CC. Why what you learned in preschool is crucial at work. The New York Times. 2015 Oct 16; Retrieved from http://www.nytimes.com/2015/10/18/upshot/how-the-modern-workplace-has-become-more-like-preschool.html?_r=0 .
• Moffitt TE, Arseneault L, Belsky D, Dickson N, Hancox RJ, Harrington H, Houts R, Poulton R, Roberts BW, Ross S. A gradient of childhood self-control predicts health, wealth, and public safety. Proceedings of the National Academy of Sciences. 2011; 108 (7):2693–2698. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• National Council of Teachers of Mathematics. Curriculum and evaluation standards for school mathematics. Reston, VA: Author; 1989. [ Google Scholar ]
• National Council of Teachers of Mathematics. Principles to actions: Ensuring mathematical success for all. Reston, VA: Author; 2014. [ Google Scholar ]
• National Governors Association Center for Best Practices. Common core state standards for mathematics. Washington, DC: Author; 2010. [ Google Scholar ]
• Papay JP. Different tests, different answers: The stability of teacher value-added estimates across outcome measures. American Educational Research Journal. 2011; 48 (1):163–193. [ Google Scholar ]
• Papay JP. Refocusing the debate: Assessing the purposes and tools of teacher evaluation. Harvard Educational Review. 2012; 82 (1):123–141. [ Google Scholar ]
• Pianta RC, Hamre BK. Conceptualization, measurement, and improvement of classroom processes: Standardized observation can leverage capacity. Educational Researcher. 2009; 38 (2):109–119. [ Google Scholar ]
• Pianta R, La Paro K, Payne C, Cox M, Bradley R. The relation of kindergarten classroom environment to teacher, family, and school characteristics and child outcomes. Elementary School Journal. 2002; 102 :225–38. [ Google Scholar ]
• Raudenbush SW, Bryk AS. Hierarchical linear models: Applications and data analysis methods. Second. Thousand Oaks, CA: Sage Publications; 2002. [ Google Scholar ]
• Rockoff JE, Speroni C. Subjective and objective evaluations of teacher effectiveness. American Economic Review. 2010:261–266. [ Google Scholar ]
• Rockoff JE, Staiger DO, Kane TJ, Taylor ES. Information and employee evaluation: evidence from a randomized intervention in public schools. American Economic Review. 2012; 102 (7):3184–3213. [ Google Scholar ]
• Ruzek EA, Domina T, Conley AM, Duncan GJ, Karabenick SA. Using value-added models to measure teacher effects on students’ motivation and achievement. The Journal of Early Adolescence. 2015; 35 (5–6):852–882. [ Google Scholar ]
• Segal C. Misbehavior, education, and labor market outcomes. Journal of the European Economic Association. 2013; 11 (4):743–779. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Šidák Z. Rectangular confidence regions for the means of multivariate normal distributions. Journal of the American Statistical Association. 1967; 62 (318):626–633. [ Google Scholar ]
• Spearman C. “General Intelligence,” objectively determined and measured. The American Journal of Psychology. 1904; 15 (2):201–292. [ Google Scholar ]
• Steinberg MP, Garrett R. Classroom composition and measured teacher performance: What do teacher observation scores really measure? Educational Evaluation and Policy Analysis. 2016; 38 (2):293–317. [ Google Scholar ]
• Tabachnick BG, Fidell LS. Using multivariate statistics. 4. New York: Harper Collins; 2001. [ Google Scholar ]
• Todd PE, Wolpin KI. On the specification and estimation of the production function for cognitive achievement. The Economic Journal. 2003; 113 (485):F3–F33. [ Google Scholar ]
• Tremblay RE, Masse B, Perron D, LeBlanc M, Schwartzman AE, Ledingham JE. Early disruptive behavior, poor school achievement, delinquent behavior, and delinquent personality: Longitudinal analyses. Journal of Consulting and Clinical Psychology. 1992; 60 (1):64. [ PubMed ] [ Google Scholar ]
• Tsukayama E, Duckworth AL, Kim B. Domain-specific impulsivity in school-age children. Developmental Science. 2013; 16 (6):879–893. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• U.S. Department of Education. A blueprint for reform: Reauthorization of the elementary and secondary education act. Washington, DC: U.S. Department of Education, Office of Planning, Evaluation and Policy Development; 2010. [ Google Scholar ]
• Usher EL, Pajares F. Sources of self-efficacy in school: Critical review of the literature and future directions. Review of Educational Research. 2008; 78 (4):751–796. [ Google Scholar ]
• West MR, Kraft MA, Finn AS, Martin RE, Duckworth AL, Gabrieli CF, Gabrieli JD. Promise and paradox: Measuring students’ non-cognitive skills and the impact of schooling. Educational Evaluation and Policy Analysis. 2016; 38 (1):148–170. [ Google Scholar ]
• Whitehurst GJ. Hard thinking on soft skills. Brookings Institute; Washington, DC: 2016. Retrieved from http://www.brookings.edu/research/reports/2016/03/24-hard-thinking-soft-skills-whitehurst . [ Google Scholar ]
• Whitehurst GJ, Chingos MM, Lindquist KM. Evaluating teachers with classroom observations: Lessons learned in four districts. Brown Center on Education Policy at the Brookings Institute; Washington, DC: 2014. Retrieved from http://www.brookings.edu/~/media/research/files/reports/2014/05/13-teacher-evaluation/evaluating-teachers-with-classroom-observations.pdf . [ Google Scholar ]
• Wigfield A, Meece JL. Math anxiety in elementary and secondary school students. Journal of Educational Psychology. 1988; 80 (2):210. [ Google Scholar ]
• Zernike K. Testing for joy and grit? Schools nationwide push to measure students’ emotional skills. The New York Times. 2016 Feb 29; Retrieved from http://www.nytimes.com/2016/03/01/us/testing-for-joy-and-grit-schools-nationwide-push-to-measure-students-emotional-skills.html?_r=0 .

The Promises and Challenges of Artificial Intelligence for Teachers: a Systematic Review of Research

• Original Paper
• Open access
• Published: 25 March 2022
• Volume 66 , pages 616–630, ( 2022 )

Cite this article

You have full access to this open access article

• Ismail Celik   ORCID: orcid.org/0000-0002-5027-8284 1 ,
• Muhterem Dindar 2 ,
• Hanni Muukkonen 1 &
• Sanna Järvelä 2  

52k Accesses

61 Citations

20 Altmetric

Explore all metrics

This study provides an overview of research on teachers’ use of artificial intelligence (AI) applications and machine learning methods to analyze teachers’ data. Our analysis showed that AI offers teachers several opportunities for improved planning (e.g., by defining students’ needs and familiarizing teachers with such needs), implementation (e.g., through immediate feedback and teacher intervention), and assessment (e.g., through automated essay scoring) of their teaching. We also found that teachers have various roles in the development of AI technology. These roles include acting as models for training AI algorithms and participating in AI development by checking the accuracy of AI automated assessment systems. Our findings further underlined several challenges in AI implementation in teaching practice, which provide guidelines for developing the field.

Similar content being viewed by others

The Role of AI Algorithms in Intelligent Learning Systems

Why Not Go All-In with Artificial Intelligence?

Current Trends in AI-Based Educational Processes—An Overview

Avoid common mistakes on your manuscript.

Introduction

Artificial intelligence (AI) has been penetrating our everyday lives in various ways such as through web search engines, mobile apps, and healthcare systems (Sánchez-Prieto et al., 2020 ). The swift advancement of AI technologies also has important implications for learning and teaching. In fact, AI-supported instruction is expected to transform education (Zawacki-Richter et al., 2019 ). Thus, considerable investments have been made to integrate AI into teaching and learning (Cope et al., 2020 ). A significant challenge in the effective integration of AI into teaching and learning, however, is the profit orientation of most current AI applications in education. AI developers know little about learning sciences and lack pedagogical knowledge for the effective implementation of AI in teaching (Luckin & Cukurova, 2019 ). Moreover, AI developers often fail to consider the expectations of AI end-users in education, that is, of teachers (Cukurova & Luckin, 2018 , Luckin & Cukurova, 2019 ). Teachers are considered among the most crucial stakeholders in AI-based teaching (Seufert et al., 2020 ), so their views, experiences, and expectations need to be considered for the successful adoption of AI in schools (Holmes et al., 2019 ). Specifically, to make AI pedagogically relevant, the advantages that it offers teachers and the challenges that teachers face in AI-based teaching need to be understood better. However, little attention has been paid to AI-based education from the perspective of teachers. Moreover, teachers’ skills in the pedagogical use of AI and the roles of teachers in the development of AI have been somehow ignored in literature (Langran et al., 2020 ; Seufert et al., 2020 ). To address these research gaps, this study explores the promises and challenges of AI in teaching practice that have been surfaced in research. Since the field of AI-based instruction is still developing, this study can contribute to the development of comprehensive AI-based instruction systems that allow teachers to participate in the design process.

Educational Use of Artificial Intelligence

There have been several waves of emerging educational technologies over the past few decades, and now, there is artificial intelligence (AI; Bonk & Wiley, 2020 ). The term artificial intelligence was first mentioned in 1956 by John McCarthy (Russel & Norvig, 2010 ). Baker and Smith ( 2019 ) pointed out that AI does not refer to a single technology but is defined as “computers [that] perform cognitive tasks, usually associated with human minds, particularly learning and problem-solving” (p. 10). AI is a general term that refers to diverse analytical methods. These methods can be classified as machine learning, neural networks, and deep learning (Aggarwal, 2018 ). Machine learning is defined as the capacity of a computer algorithm learning from the data to make decisions without being programmed (Popenici & Kerr, 2017 ). Although numerous machine learning models exist, the two most used models are supervised and unsupervised learning models (Alloghani et al., 2020 ). Supervised machine learning algorithms build a model based on the sample data (or training data), while unsupervised machine learning algorithms are created from untagged data (Alenezi & Faisal, 2020 ). In other words, the unsupervised model performs on its own to explore patterns that were formerly undetected by humans.

AI is used in education in different ways. For instance, AI is integrated into several instructional technologies such as chatbots (Clark, 2020 ), intelligent tutoring, and automated grading systems (Heffernan & Heffernan, 2014 ). These AI-based systems offer several opportunities to all stakeholders throughout the learning and instructional process (Chen et al., 2020 ). Previous research conducted on the educational use of AI presented AI’s support for student collaboration and personalization of learning experiences (Luckin et al., 2016 ), scheduling of learning activities and adaptive feedback on learning processes (Koedinger et al., 2012 ), reducing teachers’ workload in collaborative knowledge construction (Roll & Wylie, 2016 ), predicting the probability of learners dropping out of school or being admitted into school (Popenici & Kerr, 2017 ), profiling students’ backgrounds (Cohen et al., 2017 ), monitoring student progress (Gaudioso et al., 2012 ; Swiecki et al., 2019 ), and summative assessment such as automated essay scoring (Okada et al., 2019 ; Vij et al., 2020 ; Yuan et al., 2020 ). Despite these opportunities, the educational use of AI is more behind what is expected, unlike in other sectors (e.g., finance and health). To achieve successful AI implementation in education, various stakeholders, specifically, teachers, should participate in AI creation, development, and integration (Langran et al., 2020 ; Qin et al., 2020 ).

The Roles of Teachers in AI-based Education

The evolution of education towards digital education does not imply that people will need less teachers in the future (Dillenbourg, 2016 ). Instead of speculating if AI will replace teachers, understanding the advantages that AI offers teachers and how these advantages can change teachers’ roles in the classroom is more reasonable (Hrastinski et al., 2019 ). Salomon ( 1996 ) demonstrated this during the early stages of development of educational technology by pointing out the need to consider how learning occurs through and with computers. As for AI, Holstein et al. ( 2019 ) suggested that in the future, AI-based machines can help teachers perform what Dillenbourg ( 2013 ) emphasized as their orchestrator role in the learning and teaching process. For AI to be able to truly help teachers in this way, however, it must first learn effective orchestration of learning and teaching from teachers’ data. This is because effective teaching depends on teachers’ capability to implement appropriate pedagogical methods in their instruction (Tondeur et al., 2020 ), and their pedagogically meaningful and productive teaching incidents can serve as models for AI-based educational systems (Prieto et al., 2018 ). That is, the data collected from the learning setting orchestrated by teachers form the foundation of AI-based teaching. For example, the data may help researchers to understand when and how teaching is effectively progressing (Luckin & Cukurova, 2019 ; Luckin et al., 2016 ). To prove that the role of teachers in providing the data on features of effective learning is crucial for the development of AI algorithms, we investigated the kind of data collected from teachers and teachers’ roles in the creation of AI algorithms.

To effectively integrate AI-based education in schools, teachers must be empowered to implement such integration by endowing them with the requisite knowledge, skills, and attitudes (Häkkinen et al., 2017 ; Kirschner, 2015 ; Seufert et al., 2020 ). However, teachers’ AI-related skills have not yet been sufficiently defined because the potential of AI in education has not yet been fully exploited (Luckin et al., 2016 ). To explore teachers’ AI-related knowledge, skills, and attitudes, their engagement with AI-based systems within their teaching setting has to be investigated in detail (Dillenbourg, 2016 ; Seufert et al., 2020 ). Therefore, in this study, we reviewed empirical research on how teachers interacted with AI-based systems and how they participated in the development of AI-based education systems. We believe that our synthesis of empirical research on the topic will contribute to the identification of AI-related teaching skills and the effective implementation of AI-based education in schools with the support of teachers.

This study explored the perspective and roles of teachers in AI-based research through a systematic review of the latest research on the topic. Our specific research questions (RQs) are as follows:

RQ1—What was the distribution over time of the studies that examined teachers’ AI use?

RQ2—What data were collected from teachers in the studies on AI-based education?

RQ3—What were the roles of teachers in AI-based research?

RQ4—What advantages did AI offer teachers?

RQ5—What challenges did teachers face when using AI for education?

RQ6—Which AI methods were utilized in AI-based research that teachers participated in?

Table 1 below lists these RQs with their corresponding rationales.

Manuscript Search and Selection Criteria

In reviews of research, several methods are used to select the studies that will be reviewed. Studies published in important journals of a given domain are selected from databases such as ProQuest (Heitink et al., 2016 ), Education Resources Information Center (ERIC), and the Social Science Citation Index (SSCI) (Akçayır & Akçayır, 2017 ; Kucuk et al., 2013 ). For this review, we selected English-language scientific studies on teachers’ AI use that were published in journals from the Web of Science (WoS) database within the last 20 years until 14 September 2020. We used this method because the field tags (e.g., the topic and research area) of the studies were easy to access from the WoS database (Luor et al., 2008 ). We used the following search string: “artificial intelligence,” “deep learning,” “reinforcement learning,” “supervised learning,” “unsupervised learning,” “neural network,” “ANN,” “natural language processing,” “fuzzy logic,” “decision trees,” “ensemble,” “Bayesian,” “clustering,” and “regularization.” To narrow our search, we used “teacher,” “teacher education,” “teacher professional development,” “K-12,” “middle school*,” “high school*,” “elementary school*,” and “kindergarten*.” We selected the search strings based on the main concepts of AI in education in past studies and literature reviews (Baran, 2014 ; Zawacki-Richter et al., 2019 ). Figure  1 presents our study search procedure.

Flow chart for the selection of articles

In our first search, we found 751 studies. Next, we checked them to see if they met our inclusion and exclusion criteria. Our inclusion criteria were as follows: (a) empirical studies on AI in pre-service and in-service teacher education and on in-service teachers’ use of AI; (b) studies on AI applications and algorithms (e.g., personal tutors, automated scoring, personal assistant; decision trees, and artificial neural networks) for teaching or analyzing teachers’ data; and (c) studies on data collected from in-service K-12 teachers or pre-service teachers. We excluded editorials, reviews, and studies conducted at the higher education level. After we applied the criteria, 44 articles remained suitable for inclusion in this study.

Data Coding and Analysis

The publication year of the articles was noted to determine the distribution of the studies over time (RQ1). For RQ2, the following categories and category numbers were assigned to the data collected from teachers in previous AI-based research: self-report (1), video (2), interview (3), observation (4), feedback/discourse (5), grading (6), eye tracking (7), audiovisual/accelerometry (8), and log file (9). We qualitatively analyzed the content of the 44 articles to determine the advantages and challenges of AI for teachers (RQ4 and RQ5, respectively) and teachers’ roles in AI-based instruction as found in research (RQ3). We coded the studies not with the preliminary or template coding scheme, which would have unnecessarily limited them by fitting them into a pre-determined coding scheme (Şimşek & Yıldırım, 2011 ), but with the open coding process (Akçayır & Akçayır, 2017 ; Williamson, 2015 ), which followed these steps: (1) Familiarize with the whole set of articles; (2) Choose a document randomly, consider its primary meaning, and write down your thought on such meaning on the margin of the document; (3) List all your thoughts on the subject, combine similar thoughts, create three columns for key, unique, and leftover thoughts, and put each thought in the appropriate column; (4) Code the text; (5) Find the most illustrative phrases for your thoughts and turn them into categories; (6) Decide on an abbreviation for each category and alphabetize these abbreviations; (7) Incorporate the final codes and perform the initial analysis; and (8) Recode the studies if needed. To classify the AI methods (RQ6), we used previous literature reviews of AI use in diverse areas such as higher education, medicine, and business (Borges et al., 2020 ; Contreras & Vehi, 2018 ). We performed the investigator triangulation method to ensure the reliability of the coding process (Denzin, 2017 ). Accordingly, the first author coded the articles separately and then shared the codes with the second author. We negotiated disagreements by checking the code list and the relevant studies, and we updated and renamed some categories. Finally, we recoded the studies using the final code list.

Results and Discussion

Distribution of the studies.

(RQ1—What was the distribution over time of the studies that examined teachers’ AI use?)

Our analysis indicated that the first study on teachers’ AI use was published in 2004. Of the 44 studies we reviewed, 22 were published in 2018 and the following years. It has been forecasted that the usage of educational AI applications will increase (Qin et al., 2020 ; Zawacki-Richter et al., 2019 ). Such increase is implied in our finding that the publication of studies on AI-based teaching increased after 2017. Figure  2 presents the research trend on AI and teachers.

Number of articles published by year

Figure  2 further indicates that research on teachers’ AI use in education intensified in the last four years. This implies that AI-based instruction by teachers is most likely to become more common in the near future. Supporting this, our review of literature on the topics “AI” and “education” showed that studies published between 2015 and 2019 accounted for 70% of all the studies from Web of Science and Google Scholar since 2010 (Chen et al., 2020 ). The availability of AI technologies and of educational software companies to create AI-based applications is increasing rapidly all over the world (Renz & Hilbig, 2020 ). Accordingly, it seems likely that teachers’ use of AI in the teaching process will grow and more studies will be conducted on this topic.

On the other hand, there are still fewer studies on AI use in education than in other areas such as medicine and business (Borges et al., 2020 ; Luckin & Cukurova, 2019 ). The educational technology (EdTech) market is growing much more slowly than other markets with respect to the dynamics of digital transformation. One of the reasons for this is the resistance of decision-makers such as educators, teachers, and traditional textbook publishers to the use of AI (EdTechXGlobal Report, 2016 ). Considering this resistance, it can be argued that more AI research is needed to show the pedagogical uses of AI in instructional processes and to speed up the uptake of AI technologies in education.

Data Types Collected from Teachers

(RQ2—What data were collected from teachers in the studies on AI-based education?)

Self-reported data were the most common data collected from teachers in the AI-based education studies. The researchers collected self-reported data to predict teacher-related variables such as engagement, performance, and teaching quality. In these studies, machine learning algorithms were used instead of conventional regression analysis to reveal nonlinear relationships between variables of teaching practice. For instance, Wang et al. ( 2020 ) collected data from 165 early childhood teachers to better understand indicators of quality teacher–child interaction. Similarly, in Yoo and Rho ( 2020 ), teachers’ self-reported job satisfaction was predicted by a machine learning technique. In some AI studies, teacher grades of student assignments or essays were used to train AI algorithms. For example, Yuan et al. ( 2020 ), in developing an automated scoring approach, needed expert teachers’ grades to validate their AI-based scoring system. A notable finding from our review is that self-reported grades accounted for nearly 44% of all data obtained from teachers (Fig.  3 ).

In 11 of the studies that we reviewed, teachers provided more than one type of data. The data were mostly collected during or after teachers’ instruction. Our review findings highlight the crucial role of teachers in the instructional process (e.g., Huang et al., 2010 ; Lu, 2019 ; McCarthy et al., 2016 ; Pelham et al., 2020 ). For example, Schwarz et al. ( 2018 ) presented an online learning environment that uses machine learning to inform teachers about learners’ critical moments in collaborative learning by sending the teachers warnings. In their study, they observed how the teacher guided several groups at different times in a mathematics classroom. In addition to observations, they collected interview data from the teachers about the effectiveness of the online environment. Our review indicates that there is a significant gap in physiological data collection in AI studies with teachers. Only one of the studies we reviewed collected physiological data, that is, data on eye tracking and audiovisual/accelerometry data from sensors worn by the teachers (Prieto et al., 2018 ). In fact, physiological data can be considered relevant and useful for providing process-oriented, objective metrics regarding the critical moments that impact the quality of teaching or learning in an educational activity (Järvelä et al., 2021 ).

The Roles of Teachers in AI-based Research

(RQ3—What were the roles of teachers in AI-based research?)

Our findings from our open-coding analysis indicate that teachers have seven roles in AI research. These roles and their descriptions are shown in Table 2 . As seen from the table, teachers participated in AI research as models to train AI algorithms. This role was found to be the most common role of teachers in AI-based instruction ( f  = 18). This finding underlines the pivotal role of teachers in the development of AI-based education systems. For instance, Kelly et al. ( 2018 ) conducted a study to train AI algorithms to automatically detect teachers’ authentic questions in real-life classrooms. During the training of the AI algorithms, the teachers’ effective authentic questions were fed to the AI system as features. Following the AI training, the researchers tested AI in a different classroom and found that AI successfully identified authentic questions.

Another role that teachers were observed to have in AI research was providing big data to AI systems to enable them to forecast teachers’ professional development. In this line of research, teachers mostly provided data to AI systems for the latter’s prediction of different variables of the professional development of teachers such as their job satisfaction, performance, and engagement. For example, in one study, 10,642 teachers answered a survey (Buddhtha et al., 2019 ). Then, using AI, predictors of teacher engagement were determined. Similar to other areas, big data have played an important role in education, and teachers are considered among the most important sources of big data (Ruiz-Palmero et al., 2020 ). Our findings imply that AI can effectively inform teachers of their professional development.

This study also found that teachers involved in AI research provided input information on students’ characteristics for the AI-based implementation. For example, Nikiforos et al. ( 2020 ) investigated automatic detection of learners’ aggressive behavior in a virtual learning community. The AI system utilized teacher observations of students’ behavioral characteristics to predict the students who were more likely to bully others in the online community. Our review further revealed that teachers have taken on the role of grading assignments and essays to test the accuracy of AI algorithms in grading student performance. In such studies, the accuracy rate of the AI-based assessment was determined with the help of experienced teacher assessments (Bonneton-Botté et al., 2020 ; Gaudioso et al., 2012 ; McCarthy et al., 2016 ; Yuan et al., 2020 ).

In some AI-based education studies, teachers determined the criteria for some components of AI-based systems and assessments. For example, Huang et al. ( 2010 ) investigated the effect of the learning assistance tool ICT Literacy . The tool used machine learning. In their study, experienced teachers guided the AI system by defining the criteria for effective and timely feedback. In some studies, teachers also provided pedagogical guidance on the selection of materials for AI-based implementation. For example, Fitzgerald et al. ( 2015 ) utilized AI to present learning content with varying degrees of text complexity to early-grade students. They attempted to explore early-grade text complexity features. Text complexity in the AI system was determined based on teachers’ pedagogical guidance. Furthermore, teachers commented on the usability and design of AI-based technologies (Burstein et al., 2004 ). Finally, our results revealed a notable absence of pre-service teachers as participants in AI use studies. That is, there were no studies in which pre-service teachers actively participated or interacted with AI technologies.

Advantages of AI for Teachers

(RQ4—What advantages did AI offer teachers?)

We found several advantages of AI from our review of selected empirical studies on teachers’ AI use. The open coding revealed three categories of AI advantages: planning, implementation, and assessment (see Table 3 ).

The advantages of AI related to planning involved receiving information on students’ backgrounds and assisting teachers in deciding on the learning content during lesson planning. In a study, an AI system provided teachers background information on students’ risk factors for delinquency, such as aggression (Pelham et al., 2020 ). In terms of teacher assistance in planning learning content, Dalvean and Enkhbayar ( 2018 ) used machine learning to classify the readability of English fiction texts. The results of their study suggested that the classification can help English teachers to plan the course contents considering the readability features (Table 4 ).

Implementation

According to our review (see Table 3 ), the most prominent advantage of AI was stated as timely monitoring of learning processes ( f  =  12 ). For example, Su et al. ( 2014 ) developed a sensor-based learning concentration detection system using AI in a classroom environment. The system allowed teachers to monitor the degree of students’ concentration on lesson activities. Such AI-based monitoring can help teachers to provide immediate feedback (Burstein et al., 2004 ; Huang et al., 2010 , 2011 ) and quickly perform the necessary interventions (Nikiforos et al., 2020 ; Schwarz et al., 2018 ). For instance, teachers were able to discover critical moments in group learning and provide adaptive interventions for all the groups (Schwarz et al., 2018 ). Hence, AI systems can decrease the teaching burden on teachers by providing them feedback and assisting them with planning interventions and with student monitoring. In several studies, these contributions to teachers were particularly emphasized (Lu, 2019 ; Ma et al., 2020 ). Therefore, we assume that reduced teaching load may be another significant advantage of AI systems in education. For example, researchers reported that teachers benefitted from an AI-based peer tutor recommender system and saved time for other activities (Ma et al., 2020 ).

Our findings further revealed that AI can enable teachers to select or adapt the optimum learning activity based on AI feedback. For example, in Bonneton-Botté et al. ( 2020 ), teachers decided to implement exercises such as writing letters and numbers for students with a low graphomotor level based on the feedback they received from AI. According to our synthesis, AI can also make the teaching process more interesting for teachers. Teachers reported that AI-tutors facilitated enjoyable teaching experiences for them by breaking the monotony in the classroom (McCarthy et al., 2016 ). We also found out that AI algorithms can increase opportunities for teacher-student interaction by capturing and analyzing data from productive moments (Lamb & Premo, 2015 ) and tracking student progress (Farhan et al., 2018 ).

According to our review, AI helps teachers in exam automation and essay scoring and in decision-making on student performance. It has been found that an automated essay scoring system can not only significantly advance the effectiveness of essay scoring but also make scoring more objective (Yuan et al., 2020 ). Therefore, researchers are interested in the use of AI affordances to investigate automated systems. An important utility of AI-based applications in the context of assessment is to detect plagiarism in student essays (Dawson et al., 2020 ). Several existing AI-based systems (e.g., Turnitin) allow teachers to check the authenticity of essays submitted by students in graduate courses (Alharbi & Al-Hoorie, 2020 ). This can be considered an important utility of AI in student assessment. We coded seven studies on the advantage of exam automation and essay scoring. Six of these studies investigated the scoring of student-related outcomes (Annabestani et al., 2020 ; Huang et al., 2010 ; Tepperman et al., 2010 ; Yuan et al., 2020 ; Vij et al., 2020 ; Yang, 2012 ), and one study used AI-based systems to score teachers’ open-ended responses, to assess usable mathematics teaching knowledge (Kersting et al., 2014 ). We suggest that more studies be conducted on automatic scoring of teacher-related variables such as technological and pedagogical knowledge. Considering that classroom video analysis (CVA) assessment is capable of scoring and assessing teacher knowledge (Kersting et al., 2014 ), CVA can be used in both in-service and pre-service teacher education, particularly on micro-teaching methods. For example, natural language processing methods (Bywater et al., 2019 ) can utilize existing CVA scoring schemes to detect teachers’ verbal communication patterns in conveying instructional content to students. Furthermore, machine vision methods (Ozdemir & Tekin, 2016 ) can be applied to teachers’ video recordings to observe the patterns in their body posture. Such methods may provide valuable feedback to novice teachers on developing their teaching skills.

AI could also help provide teachers feedback on the effectiveness of their instructional practice (Farhan et al., 2018 ; Lamb & Premo, 2015 ). Teachers’ pedagogically meaningful teaching aspects can be modeled automatically using multiple data sources and AI (Dillenbourg, 2016 ; Prieto et al., 2018 ). Through these models, teachers can improve their instructional practices. Besides, the pedagogically effective models can train AI algorithms to make them more sophisticated.

Also, AI technologies were used to better predict or assess teacher performance or outcomes. Researchers predicted pre-service or in-service teachers’ professional development outcomes such as course achievement using machine learning algorithms, which are beneficial in revealing complex and nonlinear relationships. While seven studies collected data from in-service teachers, two studies obtained data from pre-service teachers (Akgün & Demir, 2018 ; Demir, 2015 ).

In addition, Cohen et al. ( 2017 ) conducted a study on a sample with autism spectrum disorder and another sample without. The results revealed that a machine learning tool can provide accurate and informative data for diagnosing autism spectrum disorder. In the study of Cohen et al., teachers commented on the accuracy of the tool.

Figure  4 illustrates the role of teachers in AI research and the advantages of AI for teachers. This gives us ideas about AI expectations from teachers and AI opportunities for teachers.

Advantages of AI and teacher roles in AI research

Challenges in AI Use by Teachers

(RQ5—What challenges did teachers face when using AI for education?)

The challenges in teachers’ use of AI are summarized in Table 3 . One of the most observed challenges is the limited technical capacity of AI. For example, AI may not be efficient for scoring graphics or figures and text. Fitzgerald et al. ( 2015 ) reported that an AI-based system failed to assess the complexity of texts when they included images. The limited reliability of the AI algorithm was found to be another considerable challenge. Therefore, automated writing evaluation technologies that use AI algorithms have to be improved to provide trustworthy evaluations for teachers (Qian et al., 2020 ). Inefficiency of AI systems in assessment and evaluation is related more to validity than to reliability. AI-based scoring may sometimes improperly evaluate performance (Lu, 2019 ). Our review further indicated that AI systems may be too context-dependent such that using them in varying educational settings can be challenging. For example, an AI algorithm designed to detect specific behavior in a specific online learning environment cannot work in different languages (Nikiforos et al., 2020 ). In other words, this limitation can stem from cultural differences.

The lack of technological knowledge of teachers (Chiu & Chai, 2020 ) and the lack of technical infrastructure in schools (McCarthy et al., 2016 ) are two other challenges in integrating AI into education. It has also been reported that AI-based feedback is sometimes slow. This can lead to teacher boredom in using AI (McCarthy et al., 2016 ). Although adaptive and personalized feedback is important for teachers to reduce their workload, AI systems are not always capable of giving different kinds of feedback based on students’ needs (Burstein et al., 2004 ). Therefore, AI systems currently fall short of meeting the needs of teachers for effective feedback (Fig.  5 ).

AI methods in the reviewed studies

AI Methods in Research

(RQ6—Which AI methods were utilized in AI-based research that teachers participated in?)

We coded AI methods in the studies, following previous reviews (Borges et al., 2020 ; Contreras & Vehi, 2018 ; Saa et al., 2019 ). Artificial neural networks (ANN) appeared to have been the most used ( f  =  16 ) AI method in the education studies involving teachers. ANN is a machine learning method that is widely used in business, economics, engineering, and higher education (Musso et al., 2013 ). According to our review, ANN also processes common data sourced from teachers. For example, Alzahrani and his colleagues (Alzahrani et al., 2020 ) investigated the relationship between thermal comfort and teacher performance. Through ANN analysis, they analyzed the data related to teachers’ productivity and the classroom temperature. Decision trees, another machine learning algorithm, were frequently utilized in our reviewed studies. For instance, Gaudioso et al. ( 2012 ) used decision tree algorithms on data to support teachers in detecting moments in which students were having problems in an adaptive educational system. Similar to our findings, a review of predictive machine learning methods for university students’ academic performance found that the decision tree algorithm was the most commonly used (Saa et al., 2019 ).

In our review, we also investigated the subject domains of teachers’ AI-based instruction. The studies with teachers from various domains accounted for 16% of all research (see Fig.  6 ). These studies generally had a larger sample size than the studies with teachers from a single domain (e.g., Buddhtha et al., 2019 ). Primary education and the English language appeared to be the domains where teachers use AI the most. Studies on automated essay scoring and adaptive feedback were conducted in English language courses. We found that 46% of all the studies we reviewed were performed in fields related to science, technology, engineering, and mathematics (STEM), and a much smaller percentage of studies were performed in the social science and early childhood fields together. These might have been because teachers in STEM fields are more accustomed to technology use (Chai et al., 2020 ).

Distribution of studies by subject domain

Conclusions and Future Research

Due to the growing interest in AI use, the number of studies on teachers’ use of AI has been increasing in the last few years, and more studies are needed to know more about teachers’ AI use. As AI continues to become popular in education, undoubtedly more research will focus on AI use in teachers’ instruction. Our synthesis of relevant studies shows that there has been little interest in investigating AI in pre-service teacher education. Hence, we recommend more empirical studies on pre-service teachers’ AI use. Developing AI awareness and skills among pre-service teachers may facilitate better adoption of AI-based teaching in future classrooms. As Valtonen et al. ( 2021 ) have shown, teachers’ and students’ use of emerging technologies can make a major contribution to the development of 21st-century practices in schools.

Another gap we found in our review is the limited variety of methods and data channels used in AI-based systems. It seems that AI-based systems in education do not exploit the potential of multimodal data. Most of the AI applications that teachers use utilize only self-reported and/or observation data, while different data modalities can create more opportunities to understand teaching and learning processes (Järvelä & Bannert, 2021 ). Enriching AI systems with other data types (e.g., physiological data) may give a better understanding of different layers of teaching and learning, and thus, help teachers to plan effective learning interventions, provide timely feedback and conduct more accurate assessments of students’ cognitive and emotional states during the instruction. Utilizing multimodal data can help to model more efficient and effective AI systems for education. Thus, we conclude that further work is necessary to improve the capabilities of AI systems with multimodal data.

Our review revealed that teachers have limited involvement in the development of AI-based education systems. Although in some studies, experienced teachers were recruited to train AI algorithms, further efforts are needed to involve a wider population of teachers in developing AI systems. Such involvement should go beyond training AI algorithms and involve teachers in the crucial decision-making processes on how (not) to develop AI systems for better teaching. For their part, AI developers and software companies should consider involving teachers in the development process to a greater extent.

This study showed that AI has been reported as generally beneficial to teachers’ instruction. Teachers can take advantage of AI in their planning, implementation, and assessment work. AI assists them in identifying their students’ needs so that they can determine the most suitable learning content and activities for their students. During the activities, such as a collaborative task, with the help of AI, teachers can monitor their students in a timely manner and give them immediate feedback (e.g., Swiecki et al., 2019 ). After the instruction, AI-based automated scoring systems can help teachers with assessment (e.g., Kersting et al., 2014 ). These advantages mainly reduce teachers’ workload and help them to focus their attention on critical issues such as timely intervention and assessment (Vij et al., 2020 ). However, many of the studies reviewed were conducted to predict outcome variables (e.g., performance, engagement, and job satisfaction) through machine learning algorithms (Yoo & Rho, 2020 ). More studies are needed to enable AI systems to provide information and feedback on how the learning processes temporally unfold during teachers’ instruction. Then, teachers will be able to interact with actual AI systems to better understand possible opportunities.

This study revealed several limitations and challenges of AI for teachers’ use such as its limited reliability, technical capacity, and applicability in multiple settings. Future empirical research is necessary to address the challenges reported in this study. We conclude that developing AI systems that are technically and pedagogically capable of contributing to quality education in diverse learning settings is yet to be achieved. To achieve this objective, multidisciplinary collaboration between multiple stakeholders (e.g., AI developers, pedagogical experts, teachers, and students) is crucial. We hope that this review will serve as a springboard for such collaboration.

*References marked with an asterisk show the articles used in the review

Aggarwal, C. C. (2018). Neural networks and deep learning.  Springer ,  10 , 978-3. https://doi.org/10.1007/978-3-319-94463-0

Akçayır, M., & Akçayır, G. (2017). Advantages and challenges associated with augmented reality for education: A systematic review of the literature. Educational Research Review, 20 , 1–11. https://doi.org/10.1016/j.edurev.2016.11.002

Article   Google Scholar  

*Akgün, E., & Demir, M. (2018). Modeling course achievements of elementary education teacher candidates with artificial neural networks.  International Journal of Assessment Tools in Education ,  5 (3), 491–509. https://doi.org/10.21449/ijate.444073

Alenezi, H. S., & Faisal, M. H. (2020). Utilizing crowdsourcing and machine learning in education: Literature review.  Education and Information Technologies , 1-16. https://doi.org/10.1007/s10639-020-10102-w

Alharbi, M. A., & Al-Hoorie, A. H. (2020). Turnitin peer feedback: Controversial vs. non-controversial essays. International Journal of Educational Technology in Higher Education, 17 , 1–17. https://doi.org/10.1186/s41239-020-00195-1

Alloghani, M., Al-Jumeily, D., Mustafina, J., Hussain, A., & Aljaaf, A. J. (2020). A systematic review on supervised and unsupervised machine learning algorithms for data science. In  Supervised and Unsupervised Learning for Data Science  (pp. 3–21). Springer, Cham. https://doi.org/10.1007/978-3-030-22475-2_1

Alzahrani, H., Arif, M., Kaushik, A., Goulding, J., & Heesom, D. (2020). Artificial neural network analysis of teachers’ performance against thermal comfort. International Journal of Building Pathology and Adaptation . https://doi.org/10.1108/IJBPA-11-2019-0098

Annabestani, M., Rowhanimanesh, A., Mizani, A., & Rezaei, A. (2020). Fuzzy descriptive evaluation system: Real, complete and fair evaluation of students. Soft Computing, 24 (4), 3025–3035. https://doi.org/10.1007/s00500-019-04078-0

Baker, T., & Smith, L. (2019). Educ-AI-tion rebooted? Exploring the future of artificial intelligence in schools and colleges . Retrieved from Nesta Foundation website: https://media.nesta.org.uk/documents/Future_of_AI_and_education_v5_WEB.pdf

Baran, E. (2014). A review of research on mobile learning in teacher education. Journal of Educational Technology & Society, 17 (4), 17–32.

Google Scholar  

*Bonneton-Botté, N., Fleury, S., Girard, N., Le Magadou, M., Cherbonnier, A., Renault, M., ... & Jamet, E. (2020). Can tablet apps support the learning of handwriting? An investigation of learning outcomes in kindergarten classroom.  Computers & Education ,  151 , 103831. https://doi.org/10.1016/j.compedu.2020.103831

Borges, A. F., Laurindo, F. J., Spínola, M. M., Gonçalves, R. F., & Mattos, C. A. (2020). The strategic use of artificial intelligence in the digital era: Systematic literature review and future research directions.  International Journal of Information Management , 102225. https://doi.org/10.1016/j.ijinfomgt.2020.102225

Bonk, C. J., & Wiley, D. A. (2020). Preface: Reflections on the waves of emerging learning technologies. Educational Technology Research and Development, 68 (4), 1595–1612. https://doi.org/10.1007/s11423-020-09809-x

Buddhtha, S., Natasha, C., Irwansyah, E., & Budiharto, W. (2019). Building an artificial neural network with backpropagation algorithm to determine teacher engagement based on the indonesian teacher engagement index and presenting the data in a Web-Based GIS. International Journal of Computational Intelligence Systems, 12 (2), 1575–1584. https://doi.org/10.2991/ijcis.d.191101.003

Burstein, J., Chodorow, M., & Leacock, C. (2004). Automated essay evaluation: The Criterion online writing service. Ai Magazine, 25 (3), 27–27. https://doi.org/10.1609/aimag.v25i3.1774

Bywater, J. B., Chiu J. l., Hong J., & Sankaranarayanan,V. (2019). The teacher responding tool: Scaffolding the teacher practice of responding to student ideas in mathematics classrooms.  Computers & Education   139 , 16-30. https://doi.org/10.1016/j.compedu.2019.05.004

Chai, C. S., Jong, M., & Yan, Z. (2020). Surveying Chinese teachers’ technological pedagogical STEM knowledge: A pilot validation of STEM-TPACK survey. International Journal of Mobile Learning and Organisation, 14 (2), 203–214. https://doi.org/10.1504/IJMLO.2020.106181

Chen, L., Chen, P., & Lin, Z. (2020). Artificial intelligence in education: A review. IEEE Access, 8 , 75264–75278. https://doi.org/10.1109/ACCESS.2020.2988510

Chiu, T. K., & Chai, C. S. (2020). Sustainable curriculum planning for artificial intelligence education: A self-determination theory perspective. Sustainability, 12 (14), 5568. https://doi.org/10.3390/su12145568

Clark, D. (2020).  Artificial ıntelligence for learning: How to use AI to support employee development . Kogan Page Publishers.

*Cohen, I. L., Liu, X., Hudson, M., Gillis, J., Cavalari, R. N., Romanczyk, R. G., ... & Gardner, J. M. (2017). Level 2 Screening with the PDD Behavior Inventory: Subgroup Profiles and Implications for Differential Diagnosis.  Canadian Journal of School Psychology ,  32 (3-4), 299-315. https://doi.org/10.1177/0829573517721127

Contreras, I., & Vehi, J. (2018). Artificial intelligence for diabetes management and decision support: Literature review. Journal of Medical Internet Research, 20 (5), e10775. https://doi.org/10.2196/10775

Cope, B., Kalantzis, M., & Searsmith, D. (2020). Artificial intelligence for education: Knowledge and its assessment in AI-enabled learning ecologies.  Educational Philosophy and Theory , 1–17.

Cukurova, M., & Luckin, R. (2018). Measuring the impact of emerging technologies in education: A pragmatic approach. Springer, Cham. https://discovery.ucl.ac.uk/id/eprint/10068777

*Dalvean, M., & Enkhbayar, G. (2018). Assessing the readability of fiction: a corpus analysis and readability ranking of 200 English fiction texts* 4.  Linguistic Research ,  35 , 137–170. https://doi.org/10.17250/khisli.35.201809.006

Dawson, P., Sutherland-Smith, W., & Ricksen, M. (2020). Can software improve marker accuracy at detecting contract cheating? A pilot study of the Turnitin authorship investigate alpha. Assessment & Evaluation in Higher Education, 45 (4), 473–482.

*Demir, M. (2015). Predicting pre-service classroom teachers’ civil servant recruitment examination’s educational sciences test scores using artificial neural networks.  Educational Sciences: Theory & Practice ,  15 (5). Retrieved from https://doi.org/10.12738/estp.2015.5.0018

Denzin, N. K. (2017).  The research act: A theoretical introduction to sociological methods . Transaction publishers.

Dillenbourg, P. (2013). Design for classroom orchestration. Computers & Education, 69, 485–492. https://doi.org/10.1016/j.compedu.2013.04.013 .

Dillenbourg, P. (2016). The evolution of research on digital education. International Journal of Artificial Intelligence in Education, 26 (2), 544–560. https://doi.org/10.1007/s40593-016-0106-z

EdTechXGlobal. (2016). EdTechXGlobal report 2016—Global EdTech industry report: a map for the future of education and work. Retrieved from http://ecosystem.edtechxeurope.com/2016-edtech-report

Farhan, M., Jabbar, S., Aslam, M., Ahmad, A., Iqbal, M. M., Khan, M., & Maria, M. E. A. (2018). A real-time data mining approach for interaction analytics assessment: IoT based student interaction framework. International Journal of Parallel Programming, 46 (5), 886–903. https://doi.org/10.1007/s10766-017-0553-7

Fitzgerald, J., Elmore, J., Koons, H., Hiebert, E. H., Bowen, K., Sanford-Moore, E. E., & Stenner, A. J. (2015). Important text characteristics for early-grades text complexity. Journal of Educational Psychology, 107 (1), 4. https://doi.org/10.1037/a0037289

Gaudioso, E., Montero, M., & Hernandez-Del-Olmo, F. (2012). Supporting teachers in adaptive educational systems through predictive models: A proof of concept. Expert Systems with Applications, 39 (1), 621–625. https://doi.org/10.1016/j.eswa.2011.07.052

Häkkinen, P., Järvelä, S., Mäkitalo-Siegl, K., Ahonen, A., Näykki, P., & Valtonen, T. (2017). Preparing teacher students for 21st century learning practices (PREP 21): A framework for enhancing collaborative problem solving and strategic learning skills. Teachers and Teaching: Theory and Practice, 23 (1), 25–41. https://doi.org/10.1080/13540602.2016.1203772

Heffernan, N. T., & Heffernan, C. L. (2014). The ASSISTments ecosystem: Building a platform that brings scientists and teachers together for minimally invasive research on human learning and teaching. International Journal of Artificial Intelligence in Education, 24 (4), 470–497. https://doi.org/10.1007/s40593-014-0024-x

Heitink, M. C., Van der Kleij, F. M., Veldkamp, B. P., Schildkamp, K., & Kippers, W. B. (2016). A systematic review of prerequisites for implementing assessment for learning in classroom practice. Educational Research Review, 17 , 50–62. https://doi.org/10.1016/j.edurev.2015.12.002

Holmes, W., Bialik, M., & Fadel, C. (2019). Artificial intelligence in education: Promises and Implications for Teaching and Learning . Center for Curriculum Redesign.

Holstein, K., McLaren, B. M., & Aleven, V. (2019). Co-designing a real-time classroom orchestration tool to support teacher–AI complementarity.  Journal of Learning Analytics ,  6 (2), 27–52. https://doi.org/10.18608/jla.2019.62.3

Hrastinski, S., Olofsson, A. D., Arkenback, C., Ekström, S., Ericsson, E., Fransson, G., ... & Utterberg, M. (2019). Critical imaginaries and reflections on artificial intelligence and robots in post digital K-12 education.  Post digital Science and Education ,  1 (2), 427-445. https://doi.org/10.1007/s42438-019-00046-x

*Huang, C. J., Liu, M. C., Chang, K. E., Sung, Y. T., Huang, T. H., Chen, C. H., ... & Chang, T. Y. (2010). A learning assistance tool for enhancing ICT literacy of elementary school students.  Journal of Educational Technology & Society ,  13 (3), 126-138.

Huang, C. J., Wang, Y. W., Huang, T. H., Chen, Y. C., Chen, H. M., & Chang, S. C. (2011). Performance evaluation of an online argumentation learning assistance agent. Computers & Education, 57 (1), 1270–1280. https://doi.org/10.1016/j.compedu.2011.01.013

Järvelä, S. & Bannert, M. (2021). Temporal and adaptive processes of regulated learning – What can multimodal data tell? Learning and Instruction , 72, https://doi.org/10.1016/j.learninstruc.2019.101268

Järvelä, S., Malmberg, J., Haataja, E., Sobocinski, M., & Kirschner, P. A. (2021). What multimodal data can tell us about the students’ regulation of their learning process.  Learning and Instruction ,  101203 . https://doi.org/10.1016/j.learninstruc.2019.04.004

Kelly, S., Olney, A. M., Donnelly, P., Nystrand, M., & D’Mello, S. K. (2018). Automatically measuring question authenticity in real-world classrooms. Educational Researcher, 47 (7), 451–464. https://doi.org/10.3102/0013189X18785613

Kersting, N. B., Sherin, B. L., & Stigler, J. W. (2014). Automated scoring of teachers’ open-ended responses to video prompts: Bringing the classroom-video-analysis assessment to scale. Educational and Psychological Measurement, 74 (6), 950–974. https://doi.org/10.1177/0013164414521634

Kirschner, P. A. (2015). Do we need teachers as designers of technology enhanced learning? Instructional Science, 43 (2), 309–322. https://doi.org/10.1007/s11251-015-9346-9

Koedinger, K. R., Corbett, A. T., & Perfetti, C. (2012). The Knowledge-Learning-Instruction framework: Bridging the science-practice chasm to enhance robust student learning. Cognitive Science, 36 (5), 757–798. https://doi.org/10.1111/j.1551-6709.2012.01245.x

Kucuk, S., Aydemir, M., Yildirim, G., Arpacik, O., & Goktas, Y. (2013). Educational technology research trends in Turkey from 1990 to 2011. Computers & Education, 68 , 42–50. https://doi.org/10.1016/j.compedu.2013.04.016

Lamb, R., & Premo, J. (2015). Computational modeling of teaching and learning through application of evolutionary algorithms. Computation, 3 (3), 427–443. https://doi.org/10.3390/computation3030427

Langran, E., Searson, M., Knezek, G., & Christensen, R. (2020). AI in Teacher Education. In  Society for Information Technology & Teacher Education International Conference  (pp. 735–740). Association for the Advancement of Computing in Education (AACE). https://www.learntechlib.org/p/215821/

Lu, X. (2019). An empirical study on the artificial intelligence writing evaluation system in China CET. Big Data, 7 (2), 121–129. https://doi.org/10.1089/big.2018.0151

Luckin, R., & Cukurova, M. (2019). Designing educational technologies in the age of AI: A learning sciences-driven approach. British Journal of Educational Technology, 50 (6), 2824–2838. https://doi.org/10.1111/bjet.12861

Luckin, R., Holmes, W., Griffiths, M., & Forcier, L. B. (2016). Intelligence unleashed: An argument for AI in education . Pearson Education.

Luor, T., Johanson, R. E., Lu, H. P., & Wu, L. L. (2008). Trends and lacunae for future computer assisted learning (CAL) research: An assessment of the literature in SSCI journals from 1998–2006. Journal of the American Society for Information Science and Technology, 59 (8), 1313–1320. https://doi.org/10.1002/asi.20836

Ma, Z. H., Hwang, W. Y., & Shih, T. K. (2020). Effects of a peer tutor recommender system (PTRS) with machine learning and automated assessment on vocational high school students’ computer application operating skills. Journal of Computers in Education, 7 (3), 435–462. https://doi.org/10.1007/s40692-020-00162-9

McCarthy, T., Rosenblum, L. P., Johnson, B. G., Dittel, J., & Kearns, D. M. (2016). An artificial intelligence tutor: A supplementary tool for teaching and practicing braille. Journal of Visual Impairment & Blindness, 110 (5), 309–322. https://doi.org/10.1177/0145482X1611000503

Musso, M. F., Kyndt, E., Cascallar, E. C., & Dochy, F. (2013). Predicting general academic performance and ıdentifying the differential contribution of participating variables using artificial neural networks.  Frontline Learning Research ,  1 (1), 42–71. https://doi.org/10.14786/flr.v1i1.13

Nikiforos, S., Tzanavaris, S., & Kermanidis, K. L. (2020). Virtual learning communities (VLCs) rethinking: Influence on behavior modification—bullying detection through machine learning and natural language processing. Journal of Computers in Education, 7 , 531–551. https://doi.org/10.1007/s40692-020-00166-5

Okada, A., Whitelock, D., Holmes, W., & Edwards, C. (2019). e-Authentication for online assessment: A mixed-method study. British Journal of Educational Technology, 50 (2), 861–875.

Ozdemir, O., & Tekin, A. (2016). Evaluation of the presentation skills of the pre-service teachers via fuzzy logic. Computers in Human Behavior, 61 , 288–299. https://doi.org/10.1016/j.chb.2016.03.013

Pelham, W. E., Petras, H., & Pardini, D. A. (2020). Can machine learning improve screening for targeted delinquency prevention programs? Prevention Science, 21 (2), 158–170. https://doi.org/10.1007/s11121-019-01040-2

Popenici, S. A., & Kerr, S. (2017). Exploring the impact of artificial intelligence on teaching and learning in higher education. Research and Practice in Technology Enhanced Learning, 12 (1), 1–13. https://doi.org/10.1186/s41039-017-0062-8

Prieto, L. P., Sharma, K., Kidzinski, Ł, Rodríguez-Triana, M. J., & Dillenbourg, P. (2018). Multimodal teaching analytics: Automated extraction of orchestration graphs from wearable sensor data. Journal of Computer Assisted Learning, 34 (2), 193–203. https://doi.org/10.1111/jcal.12232

Qian, L., Zhao, Y., & Cheng, Y. (2020). Evaluating China’s automated essay scoring system iWrite. Journal of Educational Computing Research, 58 (4), 771–790. https://doi.org/10.1177/0735633119881472

Qin, F., Li, K., & Yan, J. (2020). Understanding user trust in artificial intelligence-based educational systems: Evidence from China. British Journal of Educational Technology, 51 (5), 1693–1710. https://doi.org/10.1111/bjet.12994

Renz, A., & Hilbig, R. (2020). Prerequisites for artificial intelligence in further education: Identification of drivers, barriers, and business models of educational technology companies. International Journal of Educational Technology in Higher Education, 17 , 1–21. https://doi.org/10.1186/s41239-020-00193-3

Roll, I., & Wylie, R. (2016). Evolution and revolution in artificial intelligence in education. International Journal of Artificial Intelligence in Education, 26 (2), 582–599. https://doi.org/10.1007/s40593-016-0110-3

Ruiz-Palmero, J., Colomo-Magaña, E., Ríos-Ariza, J. M., & Gómez-García, M. (2020). Big data in education: Perception of training advisors on its use in the educational system. Social Sciences, 9 (4), 53. https://doi.org/10.3390/socsci9040053

Russel, S., & Norvig, P. (2010). Artificial intelligence - a modern approach . Pearson Education.

Saa, A. A., Al-Emran, M., & Shaalan, K. (2019). Factors affecting students’ performance in higher education: A systematic review of predictive data mining techniques. Technology, Knowledge and Learning, 24 (4), 567–598. https://doi.org/10.1007/s10758-019-09408-7

Salomon, G. (1996). Studying novel learning environments as patterns of change. In S. Vosiniadou, E. De Corte, R. Glaser & H. Mandl (Eds.). International Perspectives on the design of Technology Supported Learning. NJ: Lawrence Erlbaum Associates.

Swiecki, Z., Ruis, A. R., Gautam, D., Rus, V., & Williamson Shaffer, D. (2019). Understanding when students are active-in-thinking through modeling-in-context. British Journal of Educational Technology, 50 (5), 2346–2364. https://doi.org/10.1111/bjet.12869

Sánchez-Prieto, J. C., Cruz-Benito, J., Therón Sánchez, R., & García Peñalvo, F. J. (2020). Assessed by machines: Development of a TAM-based tool to measure ai-based assessment acceptance among students. International Journal of Interactive Multimedia and Artificial Intelligence, 6 (4), 80–86. https://doi.org/10.9781/ijimai.2020.11.009

Schwarz, B. B., Prusak, N., Swidan, O., Livny, A., Gal, K., & Segal, A. (2018). Orchestrating the emergence of conceptual learning: A case study in a geometry class. International Journal of Computer-Supported Collaborative Learning, 13 (2), 189–211. https://doi.org/10.1007/s11412-018-9276-z

Seufert, S., Guggemos, J., & Sailer, M. (2020). Technology-related knowledge, skills, and attitudes of pre-and in-service teachers: The current situation and emerging trends. Computers in Human Behavior, 115 , 106552. https://doi.org/10.1016/j.chb.2020.106552

Şimşek, H., & Yıldırım, A. (2011). Qualitative research methods in social sciences . Seçkin Publishing.

Su, Y. N., Hsu, C. C., Chen, H. C., Huang, K. K., & Huang, Y. M. (2014). Developing a sensor-based learning concentration detection system. Engineering Computations., 31 (2), 216–230. https://doi.org/10.1108/EC-01-2013-0010

Tepperman, J., Lee, S., Narayanan, S., & Alwan, A. (2010). A generative student model for scoring word reading skills. IEEE Transactions on Audio, Speech, and Language Processing, 19 (2), 348–360. https://doi.org/10.1109/TASL.2010.2047812

Tondeur, J., Scherer, R., Siddiq, F., & Baran, E. (2020). Enhancing pre-service teachers’ technological pedagogical content knowledge (TPACK): A mixed-method study. Educational Technology Research and Development, 68 (1), 319–343. https://doi.org/10.1007/s11423-019-09692-1

Valtonen, T., Hoang, N., Sointu, E., Näykki, P., Virtanen, A., Pöysä-Tarhonen, J., Häkkinen, P., Järvelä, S., Mäkitalo, K., & Kukkonen, J. (2021). How pre-service teachers perceive their 21st-century skills and dispositions: A longitudinal perspective. Computers in Human Behavior, 116 , 106643. https://doi.org/10.1016/j.chb.2020.106643

Vij, S., Tayal, D., & Jain, A. (2020). A machine learning approach for automated evaluation of short answers using text similarity based on WordNet graphs. Wireless Personal Communications, 111 (2), 1271–1282. https://doi.org/10.1007/s11277-019-06913-x

Wang, S., Hu, B. Y., & LoCasale-Crouch, J. (2020). Modeling the nonlinear relationship between structure and process quality features in Chinese preschool classrooms. Children and Youth Services Review, 109 , 104677. https://doi.org/10.1016/j.childyouth.2019.104677

Williamson, M. (2015). “I wasn’t reinventing the wheel, just operating the tools”: The evolution of the writing processes of online first-year composition students (unpublished doctorial dissertation) . Arizona State University.

Yang, C. H. (2012). Fuzzy fusion for attending and responding assessment system of affective teaching goals in distance learning. Expert Systems with Applications, 39 (3), 2501–2508. https://doi.org/10.1016/j.eswa.2011.08.102

*Yoo, J. E., & Rho, M. (2020). Exploration of predictors for Korean teacher job satisfaction via a machine learning technique, Group Mnet.  Frontiers in psychology ,  11 , 441. https://doi.org/10.3389/fpsyg.2020.00441

*Yuan, S., He, T., Huang, H., Hou, R., & Wang, M. (2020). Automated Chinese essay scoring based on deep learning.  CMC-Computers Materials & Continua ,  65 (1), 817–833. https://doi.org/10.32604/cmc.2020.010471

Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education–where are the educators? International Journal of Educational Technology in Higher Education, 16 (1), 39. https://doi.org/10.1186/s41239-019-0171-0

Download references

Open Access funding provided by University of Oulu including Oulu University Hospital.

Author information

Authors and affiliations.

Learning and Learning Processes Research Unit, Faculty of Education, University of Oulu, 90014, Oulu, Finland

Ismail Celik & Hanni Muukkonen

Learning and Educational Technology Research Unit, Faculty of Education, University of Oulu, 90014, Oulu, Finland

Muhterem Dindar & Sanna Järvelä

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Ismail Celik .

Ethics declarations

Human and animal rights.

There were no human participants and/or animals.

Informed Consent

This study is a literature review; therefore, no informed consent was needed.

Conflict of Interest

There are no potential conflicts of interest.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Celik, I., Dindar, M., Muukkonen, H. et al. The Promises and Challenges of Artificial Intelligence for Teachers: a Systematic Review of Research. TechTrends 66 , 616–630 (2022). https://doi.org/10.1007/s11528-022-00715-y

Download citation

Accepted : 07 March 2022

Published : 25 March 2022

Issue Date : July 2022

DOI : https://doi.org/10.1007/s11528-022-00715-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Artificial intelligence in education
• Systematic review
• Teacher professional development
• Technology integration
• Find a journal
• Publish with us
• Track your research

• NAEYC Login
• Member Profile
• Hello Community
• Accreditation Portal
• Online Learning
• Online Store

Popular Searches:   DAP ;  Coping with COVID-19 ;  E-books ;  Anti-Bias Education ;  Online Store

What is Teacher Research?

You are here

Teacher research is intentional, systematic inquiry by teachers with the goals of gaining insights into teaching and learning, becom­ing more reflective practitioners, effecting changes in the classroom or school, and improving the lives of children.... Teacher research stems from teachers' own questions and seeks practical solutions to issues in their professional lives.... The major components of teacher research are: conceptualization, in which teachers identify a significant problem or interest and determine relevant re­search questions; implementation, in which teachers collect and analyze data; and interpretation, in which teachers examine findings for meaning and take appropriate actions.... Teacher research is systematic in that teachers follow specific procedures and carefully document each step of the process. — " The Nature of Teacher Research " by Barbara Henderson, Daniel R. Meier, and Gail Perry

Teacher Research Resources

The resources below provide early childhood education professionals with tools to learn more about the teacher research process, explore accounts of teachers conducting research in their own classrooms, and connect with others in the field interested in teacher research.

Resources from  Voices of Practitioners

The Nature of Teacher Research Barbara Henderson, Daniel R. Meier, and Gail Perry

The Value of Teacher Research: Nurturing Professional and Personal Growth through Inquiry Andrew J. Stremmel

How To Do Action Research In Your Classroom: Lessons from the Teachers Network Leadership Institute Frances Rust and Christopher Clark

Resources From Other Publications

The resources listed here provide early childhood education professionals with tools to learn more about the teacher research process, explore accounts of teachers conducting research in their own classrooms, and connect with others in the field interested in teacher research.

American Educational Research Association (AERA) AERA encourages scholarly inquiry and promotes the dissemination and application of research results. It includes special interest groups (SIGs) devoted to early childhood and teacher research. Potential members can join AERA and then choose the Action Research or Teachers as Researchers SIGs (See “AR SIG, AERA” and “TR SIG, AERA” below.) AERA holds an annual conference with presentations of early childhood teacher research among many other sessions. www.aera.net

Action Research Special Interest Group, American Educational Research Association (AR SIG, AERA) This group builds community among those engaged in action research and those teaching others to do action research. It offers a blog, links to action research communities, and lists of action research books, journals, and conferences. http://sites.google.com/site/aeraarsig/

Teacher as Researcher Special Interest Group, American Educational Research Association (TAR SIG, AERA) This group consists of AERA members who are teacher educators and preK–12th grade educators; it aims to present teacher research at the AERA conference and elsewhere nationally. Early childhood teacher research is an important part of the group. http://www.aera.net/SIG126/TeacherasResearcherSIG126/tabid/11980/Default.aspx

The Center for Practitioner Research (CFPR) of the National College of Education at National-Louis University CFPR aims to affect education through collaborative scholarship contributing to knowledge, practice, advocacy, and policy in education. The website includes selected action research resources, including links to websites, book lists, conference information, and its online journal  Inquiry in Education . http://nlu.nl.edu/cfpr

Educational Action Research Educational Action Research  is an international journal concerned with exploring the dialogue between research and practice in educational settings. www.tandf.co.uk/journals/reac

Let’s Collaborate, Teacher Research from Access Excellence @ the National Health Museum This site includes useful supports for engaging in teacher research, including examples of K–12 research focused on science education. It offers information on starting a project, examples of teacher research projects, and links to online resources. www.accessexcellence.org/LC/TL/AR/

National Association of Early Childhood Teacher Educators (NAECTE) NAECTE promotes the professional growth of early childhood teacher educators and advocates for improvements to the field. NAECTE’s  Journal of Early Childhood Teacher Education  occasionally publishes teacher research articles, including a special issue focused on teacher research (Volume 31, Issue 3). NAECTE also provides ResearchNets, a forum to foster educational research with teacher research presentations. www.naecte.org

Networks: An On-line Journal for Teacher Research at the University of Wisconsin A venue for sharing reports of action research and discussion on inquiry for teachers at all levels, this journal provides space for discussion of inquiry as a tool to learn about practice and improve its effectiveness. http://journals.library.wisc.edu/index.php/networks

Self-Study Teacher Research: Improving your Practice through Collaborative Inquiry, Student Study Guide from Sage Publications This web-based student study site accompanies a book of the same name; it provides a wealth of information on its own for teachers or teacher educators who conduct studies of their own teaching practice. http://www.sagepub.com/samaras/default.htm

Teacher Action Research from George Mason University This site offers information about the teacher research process, including resources for carrying out teacher research studies. It also contains discussion of current teacher research issues and a comparison of teacher research to other forms of educational research and professional development. http://gse.gmu.edu/research/tr

Teacher Inquiry Communities Network from the National Writing Project (NWP) This network offers information on a mini-grant program supporting an inquiry stance toward teaching and learning. It includes information about the grant program, program reports, and examples of projects (including early elementary projects). http://www.nwp.org/cs/public/print/programs/tic

Teaching and Teacher Education This journal aims to enhance theory, research, and practice in teaching and teacher education through the publication of primary research and review papers. http://www.journals.elsevier.com/teaching-and-teacher-education

Voices of Practitioners

ORIGINAL RESEARCH article

Unfair teachers, unhappy students: longitudinal associations of perceived teacher relational unfairness with adolescent peer aggression and school satisfaction.

• Department of Developmental Psychology and Socialization, University of Padua, Padua, Italy

Introduction: Teacher relational unfairness is a significant risk factor for students’ physical and mental well-being, especially during adolescence. However, school psychology research has not yet fully analyzed the links between teacher unfairness and important indicators of school experience and wellbeing, including peer aggression and school satisfaction. Even less evidence does exist with longitudinal, multilevel data.

Methods: The present study tested the prospective relations between Fall perceived teacher unfairness and Spring reactive and proactive aggression, and school satisfaction. At T1, participants were 1,299 students (48.3% girls, mean age = 13.6 years, SD = 1.1) attending 67 classrooms in Italian public schools, whereas 1,227 students participated in the second wave 6 months later.

Results: Multilevel regressions showed that, at the individual level, T1 perceived teacher unfairness positively predicted T2 reactive and proactive aggression, and negatively predicted school satisfaction. At the class-level, T1 class teacher unfairness explained between-class variability in T2 school satisfaction, but not variability in peer aggression.

Discussion: The findings expand current knowledge about the role of teacher unfairness with the classroom and have implications for interventions at school.

1 Introduction

Teacher unfairness refers to students’ perceptions of being treated unfairly by their teachers. Relational (or interpersonal) unfairness, specifically, pertains to the evaluation of how individuals are treated in terms of fairness, honesty, and respect within their interpersonal relationships ( Colquitt, 2001 ; Lenzi et al., 2013 ; Rasooli et al., 2019 ). This represents a significant, yet relatively understudied component of negative teacher-student relationships within the classroom microsystem, which can be a potentially anxiety-inducing situation able to influence students’ behavior, well-being, and adjustment ( Bronfenbrenner, 1979 ; Swearer and Hymel, 2015 ).

When individuals perceive fairness in their treatment, they tend to view those in authority as more reliable and trustworthy; moreover, they experience an enhanced sense of self-worth and have a greater feeling of belonging and self-esteem ( Tyler and Smith, 1999 ; Cropanzano et al., 2001 ). In the classroom, teacher fairness and respect for students contribute to a better relational climate and reduce negative behaviors ( Murdock, 1999 ). Conversely, unfair treatment more likely leads to emotions such as anger, frustration, or anxiety ( Roeser et al., 1998 ).

With few exceptions, most of the knowledge about the relation of unfair treatment with individual adjustment derives from social/organizational psychology and adult samples. However, in the last decade, consistent calls for explicit investigations of fairness issues within classroom contexts have been advanced ( Kazemi, 2016 ; Sabbagh and Resh, 2016 ; Rasooli et al., 2019 ). While school-based research has focused more on distributive and procedural fairness, especially in relation with academic motivation, engagement, and performance ( Rasooli et al., 2019 ), interpersonal (or relational) fairness has been much less analyzed. In fact, a few studies so far have addressed the association between teacher relational unfairness and physical/mental health outcomes among adolescents in the school context (e.g., Santinello et al., 2009 ; Gini et al., 2018 ). Even less studies have tried to answer whether higher levels of teacher unfairness might explain adolescent students’ poor school adjustment, in terms of peer aggression and school satisfaction. Adopting a longitudinal, multilevel approach, the current study aimed at investigating this research question over the course of one school year. Specifically, it was tested whether perceived teacher unfairness in the Fall contributed to explain students’ reactive and proactive aggression and school satisfaction in the Spring, after controlling for the stability of the outcomes and for school stress related to academic demands. Moreover, beyond individual-level effects, it was analyzed the potential role of teacher unfairness at the class-level, investigating whether, on average, higher levels of teacher unfairness corresponded to increased peer aggression and reduced school satisfaction in school classes. The focus on both levels is an important novelty of this study. The study involved adolescent students because research suggested that being treated unfairly by teachers is a more frequent school stressor among adolescents compared to children ( Hjern et al., 2008 ) and because unfair treatment can be particularly destructive during early/middle adolescence due to adolescents’ heightened sensitivity to social comparisons ( Osterman, 2000 ).

1.1 Perceived teacher unfairness and peer aggression

One potential negative correlate of perceived teacher unfairness relevant for students’ school adjustment is aggressive behavior (e.g., Vieno et al., 2011 ; James et al., 2015 ). Even though many components of the classroom context might play a role in students’ aggressive behavior, one aspect that has received limited attention is how students perceive the fairness (or lack thereof) in their treatment by teachers. According to classic equity and social exchange theories (e.g., Adams, 1965 ) and the cognitive appraisal model of stress ( Lazarus, 1966 ), the social/organizational literature about justice has reported that adults’ actions of relational unfairness can stimulate anger and aggressive behavior (e.g., Skarlicki and Folger, 1997 ). Moreover, consistent with the ecological systems theory ( Bronfenbrenner, 1979 ), regular experiences of perceiving unfair treatment from teachers could potentially contribute to the propagation of social norms that tolerate disrespectful and dominating behaviors. Adolescents may view these behaviors as acceptable in the classroom context and adopt them accordingly. Ultimately, this could lead to imbalanced peer interactions characterized by dominance and aggression, or to the use of aggressive behavior as a means to address conflicting or frustrating situations among classmates. Furthermore, teachers’ unfair treatment can erode their authority legitimacy ( Tyler and Lind, 1992 ), increasing the likelihood of student involvement in aggressive behavior as they imagine they will not face consequences ( Santinello et al., 2011 ; Vieno et al., 2011 ).

According to a recent meta-analysis, “a particularly damaging link exists between teachers’ poor relationships with students and negative interactions in the peer context, at least as it concerns involvement in peer aggression and bullying” ( Krause and Smith, 2022 , p. 321). However, there have been limited studies that have specifically examined the impact of perceived teacher unfairness; these studies focused, mainly, on school bullying and employed cross-sectional designs. For example, in a large sample of early, Santinello et al. (2011) found that teacher unfairness was significantly associated with being a bully or a bully-victim, even after adjusting for several potential confounding factors, including age, sex, socio-economic status, empowerment with friends, school achievement, and trust in people. Another study ( Lenzi et al., 2014 ) found that the association between perceived teacher unfairness and school bullying was partially mediated by endorsement of instrumental goals. According to the social information processing model of aggression proposed by Crick and Dodge (1994) , this finding aligns with the notion that perceived unfairness from teachers might serve as a potential mechanism affecting students’ socio-moral cognition and promoting the occurrence of bullying behaviors. As argued by Arsenio and Gold (2006) , socio-moral cognitions, including the bias of valuing instrumental goals more than relational ones, may partly stem from unfairness experienced in different social contexts, including the classroom. Adolescents who believe they are being treated unjustly by their teachers can potentially cultivate a cynical and pessimistic perspective on the concept of morality as a form of authority. This perception may subsequently influence their behavior when interacting with their peers.

However, bullying is not the only form of peer aggression at school and research on the role of teacher unfairness in broader adolescents’ aggression conducts is necessary, especially with longitudinal data. Bullying is the most common form of proactive aggression, while many students also rely on reactive aggression to deal with peer conflicts ( Little et al., 2003 ). These two forms of aggression certainly overlap to some degrees, but they are also conceptually distinct and can have different correlates ( Polman et al., 2007 ). Little research, however, has explicitly focused on the influence of teacher unfairness for reactive and proactive aggression separately. To the best of our knowledge, only one study conducted with a large sample of Chinese adolescents ( Ren et al., 2023 ) has provided preliminary results on the association between perceived teacher unfairness and reactive and proactive aggression. Even though it was not the main aim of that study, the authors found positive bivariate correlations between teacher unfairness and both reactive and proactive aggression, both concurrently and longitudinally after 6 months. In line with this, the first aim of the present study was to test the longitudinal relations between individual student’ perceived teacher unfairness and both reactive and proactive aggression in a sample of adolescents.

1.2 Perceived teacher unfairness and school satisfaction

Teacher unfairness toward students can also influence their motivation, satisfaction with, and adjustment at school ( Stipek et al., 1998 ; Ripski and Gregory, 2009 ; Bayram Özdemir and Özdemir, 2020 ). For example, it has been shown that perceived unfair treatment is associated with lower adolescents’ academic motivation and perceived academic value ( Roeser et al., 1998 ), whereas teacher fairness is linked to students’ school engagement ( Danielsen et al., 2010 ). Students also tend to be more satisfied when they belong to classrooms where teachers treat them fairly ( Samdal et al., 1998 ). Low school satisfaction, indeed, seems to be one of the most important negative correlates of teacher unfairness. For instance, in a recent investigation with adolescents, Gini et al. (2018) have found that perceived teacher unfairness was associated with poorer psychological and physical health, lower satisfaction with school and friends at school, and lower sense of safety. Among these effects, the relation between teacher unfairness and school satisfaction was the strongest one. Interestingly, perceived teacher unfairness uniquely contributed to explain adolescent’ school satisfaction even after controlling for another important school stressor, that is, peer victimization ( Gini et al., 2018 ). However, because prior studies have employed cross-sectional designs, we aimed to expand previous findings by testing whether the negative role of perceived teacher unfairness in explaining students’ school satisfaction is confirmed longitudinally, over the course of 6 months. Moreover, we wanted to make sure that students reported lower school satisfaction was not due to other related academic reasons. Therefore, we controlled for the potential effect of school stress, in terms, for example, of high pace of schoolwork ( Hjern et al., 2008 ).

1.3 Sex differences

Literature findings on sex differences vary according to the specific variable we take into consideration. Sex differences consistently emerge for mean levels of peer aggression, with males reporting higher levels of aggressive behavior than females ( Card et al., 2008 ); to a lesser extent, sex seems to play a role also in school satisfaction, even though results do not consistently favor males or females ( Löfstedt et al., 2020 ). To the best of our understanding, only two research studies have specifically examined gender disparities when it comes to the impact of perceived teacher unfairness on individual outcomes. One study ( Lenzi et al., 2014 ) did not report significant sex differences in the cross-sectional association between perceived teacher unfairness and bullying. In the second study ( Gini et al., 2018 ), adolescent girls showed stronger links between perceived teacher unfairness and satisfaction with school. In the current work we explored whether links between the constructs of this study were different for males and females.

1.4 Perceived teacher unfairness at the class-level

While research at the class-level on broader concepts of negative teacher-student relationships—which sometimes include, but are not limited to perceived teacher unfairness—exists (e.g., Thornberg et al., 2018 ; Ten Bokkel et al., 2023 ), most of the literature focused on relational unfairness has restricted the analysis to the individual students’ perceptions and how they relate to the outcomes of interest. However, when we take into account the classroom setting, we come across a significant concern regarding fairness, specifically the environment in which a judgment is formed. In the social/organizational literature, for example, fairness context—the average of individual perceptions of fairness within a group—has been found to predict satisfaction above and beyond individual-level perceptions of fairness ( Mossholder et al., 1998 ; Naumann and Bennett, 2000 ). The given evidence indicates that certain attributes or traits of the group, which can be considered as a representation of the teacher’s fairness, might have a connection to the variations in overall satisfaction among different groups ( Wendorf and Alexander, 2005 ).

Apart from a few instances (e.g., Vieno et al., 2011 ), there has been limited exploration into the impact of class-level perceptions of relational fairness on students’ behavior or level of satisfaction. Another important limitation of the current literature of teacher relational unfairness is therefore lack of systematic investigations of both individual-level and class-level longitudinal effects. In this context, it is crucial to grasp the distinctive traits of the Italian educational system and their potential influence on how adolescents perceive unfair treatment by teachers. Students stay in the same classroom alongside a fixed group of classmates and the same teachers throughout the entire school year (and usually for more than 1 year). In the context of a classroom, the interactions between teachers and students play a significant role in shaping the overall environment. These interactions are particularly crucial and influential in comparison to other countries. Hence, an essential objective of this study was to explore the connection between unfair treatment experienced in the classroom and the subsequent manifestation of aggressive behavior and satisfaction levels of adolescents in school.

1.5 The current study

The objective of this study was to examine the relationship between fairness exhibited by teachers at both the individual and class levels and peer aggression and school satisfaction among a group of adolescents throughout a school year. At the individual level, it was hypothesized that, after controlling for the stability of the same behavior, for the other outcomes, and for academic school stress, perceived teacher unfairness measured in the Fall would be positively associated with peer aggression in the Spring. In order to contribute to the current body of knowledge, this study incorporated both reactive and proactive aggression as potential outcomes. Due to lack of previous data, it is uncertain whether perceived teacher unfairness might be a stronger risk factor for one type of aggression than another, or whether it is a similar risk factor for all types of aggressive behavior. Drawing upon the restricted empirical evidence at hand, and theories of unfairness described above, it was expected to find comparable effects of perceived teacher unfairness on both forms of aggression over a period of 6 months. Moreover, we expected higher levels of perceived teacher unfairness to be associated with lower school satisfaction 6 months later. These hypotheses were based both on the theories and the previous cross-sectional findings reviewed above.

Moreover, even though a thorough analysis of sex differences was not within the scope of the current study, in the current sample we explored whether links between the constructs of this study were different for males and females. That is, we tested for a possible moderation effect of sex on the longitudinal associations between perceived teacher unfairness and the three outcomes.

Regarding the class-level analysis, a positive longitudinal association was hypothesized between class teacher unfairness and both reactive and proactive aggression. It was anticipated that there would be a higher likelihood of aggressive behavior in the Spring in school classrooms where teachers were perceived to be more unfair during the preceding Fall. Similarly, we expected that between-class variability of school satisfaction would be significantly explained by class levels of teacher unfairness, so that students would report on average lower school satisfaction if they belong to classrooms with higher levels of perceived teacher unfairness. At both levels of analysis, the hypothesized effect of perceived teacher unfairness was tested controlling for the individual-level and class-level stability of the outcomes, and for school stress.

Finally, it was examined whether there was between-class variability in the associations between perceived teacher unfairness and T2 outcomes, and whether any variation could be explained by class teacher unfairness (i.e., cross-level interaction).

2 Materials and methods

The data used in this study were extracted from a larger dataset of a longitudinal project examining the social-cognitive factors associated with aggressive behavior in adolescents. A portion of this dataset was previously utilized in another study on the moral predictors of aggressive behavior ( Gini et al., 2022 ). Although the sample is the same, there is minimal overlap between the data used in the previous study and the current study, with only age, sex, and reactive and proactive aggression being common variables, serving different research purposes. The data on teacher unfairness, school satisfaction, and school stress have never been used in previously published manuscript. A total of 67 classes from 9 public schools in urban and suburban areas of Northern Italy participated, comprising students in grades 7th to 10th (typically aged 12 when entering grade 7th in Italy). The average class size was 20.1 students.

The first wave of data collection occurred in December 2017, approximately 3 months after the beginning of the school year. At that time, 1,299 students (48.3% girls, mean age = 13.6 years, SD  = 1.1) completed the study measures. For the second wave (May 2018, near the end of the school year), 1,227 students (48.7% girls) participated, resulting in a retention rate of 94%. Out of these participants, 6 students did not respond to the reactive-proactive aggression scale, and 3 students did not complete the items about school satisfaction. Attrition analyses were conducted to examine differences between students who participated in both waves and those who did not. Findings indicated no differential attrition based on gender ( χ 2  = 0.70, p  = 0.71) or no significant differences in reactive and proactive aggression and school stress. However, students who did not participate in the second wave were slightly older than those who took part in both waves ( M age  = 14.18 vs. M age  = 13.60 ; t  = 5.71, p  < 0.001); they also reported lower school satisfaction ( M  = 2.93 vs. M  = 3.17; t  = 3.88, p  < 0.001) and higher perceived teacher unfairness ( M  = 2.41 vs. M  = 2.18; t  = 3.15, p  < 0.001).

Regarding ethnic/cultural background, the majority of participants (88.9%) had both parents born in Italy, aligning with national statistics on the Italian student population ( MIUR, 2019 ). Conversely, 11.1% of students had one or both parents born in foreign countries. Socioeconomic background was assessed using the Family Affluence Scale III ( Torsheim et al., 2016 ), a validated measure of family socioeconomic status (SES). Most participants came from medium- and high-class families (low FAS: 7.2%; medium FAS: 59.7%; high FAS: 33.1%).

2.2 Measures

2.2.1 perceived teacher unfairness.

To assess perceived teacher unfairness, a 6-item scale was employed ( Gini et al., 2018 ). This scale measured students’ perceptions regarding the extent to which they were treated fairly and respectfully by their teachers. Sample items included “My teachers treat me fairly” (reverse scored) and “I am treated too severely by my teachers.” Participants indicated their agreement on a 5-point scale, ranging from 1 (completely disagree) to 5 (completely agree). Before computing participants’ scores, positively worded items were reverse scored to ensure that higher scores reflected greater perceived teacher unfairness. Previous studies involving Italian adolescents ( Gini et al., 2018 ) have demonstrated good psychometric properties, including good test–retest reliability ( r  = 0.67). In this sample, the scale confirmed a good factorial structure (CFA: χ 2  = 2.10, p  = 0.91, CFI = 1, RMSEA = 0.00, SRMR =0.004). The internal consistency for the current sample was Cronbach’s α  = 0.76 (95% CI = 0.74–0.78), McDonald’s ω = 0.73.

2.2.2 Reactive and proactive aggression (T1 and T2)

At both waves of the study, participants’ reactive and proactive aggressive behavior was measured using a 24-item scale ( Little et al., 2003 ). Reactive aggression was assessed with 12 items describing reactions to being hurt or upset by others (e.g., “When I’m hurt by someone, I often fight back;” “If others upset or hurt me, I often tell my friends to stop liking them”), while proactive aggression was assessed with 12 items measuring aggressive actions taken to achieve personal goals (e.g., “I often start fights to get what I want,” “I often tell my friends to stop liking someone to get what I want”). Participants rated the frequency of their aggressive behavior on a 6-point scale ranging from 1 (not at all) to 6 (very much).

This scale has been previously utilized with Italian adolescents and has demonstrated good psychometric properties (e.g., Gini et al., 2015 ). Longitudinal scalar invariance in this sample was confirmed as reported in more details in the Results section. Accordingly, for each participant, responses to relevant items were averaged to obtain scores for reactive aggression and proactive aggression at T1 (reactive aggression: Cronbach’s α = 0.89, 95% CI = 0.98–0.90, McDonald’s ω = 0.93; proactive aggression: α = 0.95, 95% CI = 0.94–0.95, ω = 0.97) and T2 (reactive aggression: α = 0.89, 95% CI = 0.88–0.90, ω = 0.93; proactive aggression: α = 0.96, 95% CI = 0.95–0.96, ω = 0.98).

2.2.3 School satisfaction (T1 and T2)

Participants’ school satisfaction was measured using a subscale of the Multidimensional Students’ Life Satisfaction Scale ( Huebner, 1994 ; Gilman et al., 2000 ). This subscale consisted of 8 items measuring adolescents’ satisfaction specifically related to school (e.g., “I look forward to going to school,” “I like being in school”). At both waves, participants provided answers on a scale ranging from 1 (completely disagree) to 5 (completely agree). Negatively keyed items were reversed scored ensuring that higher scores indicated greater levels of satisfaction. Previous studies involving Italian adolescents ( Gini et al., 2018 ) have demonstrated good psychometric properties, including satisfactory test–retest reliability ( r  = 0.77). Longitudinal scalar invariance in this sample was confirmed as reported in more details in the Results section. The internal consistency of the scores in this sample was α  = 0.72 (95% CI = 0.70–0.75), ω  = 0.84 at T1 and α  = 0.74 (95% CI = 0.72–0.76), ω  = 0.85 at T2.

2.2.4 School stress

Students’ feelings of stress at school were measured using a 4-item scale adapted from previous studies ( Byrne et al., 2007 ; Hjern et al., 2008 ). Participants were asked to rate how frequently in the last 3 months they thought that, for example, the pace of the schoolwork was too high or that there were too many class assignments and oral tests. Answers were provided on a 5-point scale, ranging from 1 (never) to 5 (10 times or more). The scale showed good factorial structure (CFA: χ 2  = 3.94, p  = 0.19, CFI = 0.99, RMSEA = 0.022, SRMR = 0.008). The internal consistency for the current sample was α  = 0.77 (95% CI = 0.75–0.77, ω  = 0.78).

2.2.5 Class-level variables

To test one of our main hypothesis about the role of perceived teacher unfairness at the class-level and to account for the stability of the outcomes within each classroom, aggregated scores of each variable were created by averaging the individual scores among classmates, aligning with previous research on class norms and characteristics (e.g., Salmivalli and Voeten, 2004 ; Busching and Krahé, 2020 ; Szumski et al., 2020 ).

2.3 Procedure

First, school principals granted authorization for the classes to participate in the study. Parents of the students then provided active consent by signing a letter that informed them about the study and its objectives. Less than 10% of students in the participating classrooms did not receive parental consent. Assent for participation was also obtained from adolescents with parental consent; no one refused to participate. Data collection occurred twice within one school year, where participants completed a web-based questionnaire during a regular school hour. An anonymized alphanumeric code was used to match T1 and T2 data. A graduate research assistant was present during data collection and assured participants that their responses would remain confidential. Participants were encouraged to seek assistance if needed. At the end of data collection, any questions regarding the questionnaires or the overall aims of the project were addressed. The study protocol was approved by the local Ethics Committee for Research in Psychology (protocol #1157/2012).

2.4 Data analyses

Missing data were minimal, with only a small number of students having failed to complete the full list of items. To handle missing data, full information maximum likelihood (FIML) estimation ( Enders and Bandalos, 2001 ) in Mplus was used, so that all available information was used in the model estimation.

As a preliminary step, longitudinal confirmatory factor analyses were conducted on the aggression scores and on school satisfaction at both waves to check for longitudinal invariance. A three-factor model (reactive aggression, proactive aggression, and school satisfaction) was tested. The assumption of invariance was evaluated based on change in value of fit indices (i.e., ΔCFI, ΔRMSEA, and ΔSRMR). Negligible change, that is, a ΔCFI smaller than 0.01 and a change smaller than 0.015 in RMSEA and SRMR, was considered indicative of invariance (e.g., Cheung and Rensvold, 2002 ; Chen, 2007 ). Subsequently, multivariate multilevel modeling was performed in Mplus 8.3 ( Muthén and Muthén, 1998–2017 ). The three simultaneous dependent variables were reactive and proactive aggression and school satisfaction in Spring (T2). In this way, we took into account the intercorrelation between the outcomes, while testing the relative strength of each of the respective predictors. At the individual-level, sex (0 = males, 1 = females), age, fall levels of reactive and proactive aggression, of school satisfaction, and of school stress were entered as control variables. Perceived teacher unfairness at T1 was the key predictor. To enhance interpretability, individual-level variables, except sex, were group-mean centered. In addition to main effects, the two-way interaction between perceived teacher unfairness and sex was entered. At level 2, grade, T1 class-level aggression scores, and class-level school satisfaction and school stress were entered as control variables; T1 class-level perceived teacher unfairness was entered as main contextual predictor. Variables at the class-level were grand-mean centered.

Finally, it was examined whether there was between-classroom variability in the associations between perceived teacher unfairness and T2 aggression and school satisfaction (i.e., random slopes). In case a significant random slope emerged, cross-level interaction would be tested to check if a level-2 variable could explain the slope variability.

3.1 Descriptive statistics and correlations

Tables 1 , 2 report descriptive statistics and correlations at the individual-level and the class-level, respectively. As expected, both types of aggressive behavior and school satisfaction showed significant stability across the school year. Moreover, perceived teacher unfairness was positively, but weakly associated with both types of aggression, and negatively and strongly associated with school satisfaction, both at the individual and the class-level.

Table 1 . Means, standard deviations and correlations among variables at the individual level.

Table 2 . Means, standard deviations and correlations among variables at the class level.

3.2 Multilevel analyses

As preliminary analysis, we performed longitudinal confirmatory factor analyses on the aggression scores and on school satisfaction at both waves. First, test of configural invariance in the two waves yielded an acceptable fit: ( χ 2  = 12,064, p  < 0.001, CFI = 0.914, RMSEA = 0.062, SRMR =0.08). Second, constraining all loadings to equality across waves (i.e., metric invariance) did not lead to reduction in model fit (ΔCFI = −0.008; ΔRMSEA = −0.003; ΔSRMR = 0.000). Finally, scalar invariance was also checked (ΔCFI = 0.001; ΔRMSEA = −0.003; ΔSRMR = 0.000), confirming longitudinal invariance of the three scores.

We then estimated an unconditional model to calculate how much variance of T2 reactive and proactive aggression and T2 school satisfaction existed at the individual- and class-level. Within- and between-level variance estimates were the following: 0.685 and 0.047 for reactive aggression, 0.591 and 0.078 for proactive aggression, and 0.329 and 0.106 for school satisfaction. The intraclass correlation coefficients (ICC) therefore indicated that 6.3% of the variation of reactive aggression, 11.6% of the variation of proactive aggression, and 23.9% of the variation of school satisfaction were due to differences between classes. Based on the ICC values and an average classroom size of 20, estimated design effects were 2.20 for reactive aggression, 3.20 for proactive aggression, and 5.54 for school satisfaction, further supporting the appropriateness of adopting a multilevel analytical framework. Moreover, the ICC for teacher unfairness was 0.31, also confirming the non-independence of the unfairness perceptions within the classes and the meaningfulness of computing a class-level teacher unfairness score.

Results of the full multilevel model are reported in Table 3 . At the individual level, the model explained 34.6% of variance for reactive aggression, 27% of variance for proactive aggression and 36.3% of variance for school satisfaction. At the class level, the variance explained was 61.2% for reactive aggression, 40% for proactive aggression, and 90.3% for school satisfaction. For the sake of simplicity, the findings for the three outcome variables are summarized separately in the following paragraphs.

Table 3 . Multivariate multilevel modeling predicting T2 aggression and school satisfaction.

3.2.1 Reactive aggression

At the individual level, reactive aggression was found to be moderately stable over the course of the school year ( b  = 0.527, SE  = 0.04, p  < 0.001) and, as expected, more frequent among males ( b  = −0.213, SE  = 0.05, p  < 0.001). After taking all the other variables into account, as hypothesized, higher levels of perceived teacher unfairness at T1 positively predicted T2 students’ reactive aggression ( b  = 0.179, SE  = 0.07, p  = 0.010). At the class-level, after controlling for the stability of the aggressive behavior, no significant predictor emerged.

3.2.2 Proactive aggression

Regarding proactive aggression, the analysis at the individual level yielded one significant main effect and an interaction, beyond the expected effects of T1 proactive aggression ( b  = 0.332, SE  = 0.06, p  < 0.001) and sex ( b  = −0.285, SE  = 0.06, p  < 0.001). Consistent with our hypothesis, students who perceived higher teacher unfairness at T1 reported more proactive aggression at T2 ( b  = 0.234, SE  = 0.08, p  = 0.002). This main effect was qualified by a significant interaction between perceived teacher unfairness and sex. Simple slope analysis revealed a positive association between perceived teacher unfairness and T2 proactive aggression for male adolescents ( b  = 0.242, p  = 0.002), but not for female adolescents ( b  = 0.012, p  = 0.82).

At the class level, similar to the findings for reactive aggression, none of the predictors significantly explained between-class variation in scores of proactive aggression at the end of the school year.

3.2.3 School satisfaction

Results about school satisfaction showed that reporting more reactive aggression ( b  = −0.078, SE  = 0.03, p  = 0.006) and more school stress ( b  = −0.03, SE  = 0.01, p  = 0.034) at T1 was associated with lower school satisfaction at the end of the school year. Regarding our study hypothesis, perceived teacher unfairness was a significant negative predictor of school satisfaction ( b  = −0.129, SE  = 0.03, p  < 0.001). At the class-level, between-class variability in school satisfaction was significantly explained by class-level teacher unfairness ( b  = −0.206, SE  = 0.07, p  = 0.003), even after controlling for the stability of class-level school satisfaction and for school stress.

3.2.4 Cross-level interactions

As a final step, it was explored whether the slopes of the associations between perceived teacher unfairness and T2 aggressive behavior and school satisfaction varied across classrooms. However, no significant cross-level interactions emerged. Therefore, in the interest of parsimony, cross-level interactions were not included in the final model.

4 Discussion

As educators within the classroom, influential individuals, and catalysts for social development, teachers can shape how students interact with each other and adjust to school life. Within the broader literature on teacher-student relationships and classroom justice, the current study contributed to expand the still limited empirical evidence about teacher relational unfairness by testing its longitudinal associations with important indicators of students’ school adjustment and well-being, namely peer aggression and school satisfaction. Specifically, we included two types of aggressive behavior (i.e., reactive and proactive aggression), to further add to the current literature on teacher relational unfairness that has almost only focused on a specific subtype of aggressive behavior (i.e., bullying). The results have confirmed that when students perceive unfair treatment from their teachers, it can heighten the likelihood of them displaying aggressive behavior within the classroom. Additionally, this perception of unfair treatment can also result in lower levels of satisfaction among students with regard to their overall school experience.

4.1 Individual-level effects

First, findings from the multivariate multilevel modeling confirmed the hypothesis that perceived teacher unfairness at the individual-level significantly predicted the three outcomes of interest 6 months later. This result was robust after controlling for the stability of each outcome, for the other outcomes, and for academic school stress. In simple terms, students who reported higher levels of teacher relational unfairness in the Fall were more likely to report both reactive and proactive aggression in the Spring. Moreover, the higher the perception of being treated unfairly by teachers, the lower the school satisfaction at the end of the school year. These findings are consistent with the previous evidence reviewed above about the cross-sectional association of perceived teacher unfairness especially with bullying ( Santinello et al., 2011 ; Lenzi et al., 2014 ), fighting ( Vieno et al., 2011 ), and school satisfaction ( Samdal et al., 1998 ; Gini et al., 2018 ) that guided our study hypotheses. Moreover, they expand the current knowledge in that the negative role of teacher unfairness emerged in a short-term longitudinal design, with data modeled at both the individual- and class-level, and after controlling for potential confounders. Although not conclusive, these findings are clearly important and call for further longitudinal investigations on the role of teacher relational unfairness in other important domains of students’ school life.

Theoretically, the current findings are consistent with, and give further support to, the predictions of the social ecological theory ( Bronfenbrenner, 1979 ) and social-cognitive models of stress applied to the school context ( Swearer and Hymel, 2015 ), which indicate perceiving unfairness from teachers within the classroom microsystem as a meaningful, negative experience able to influence adolescents’ well-being and school/social adjustment. Teachers have a significant role in managing the dynamics within the classroom when it comes to their interactions with students. They not only promote and reinforce positive conduct but also discourage any negative behavior. Moreover, teachers act as a mediator, fostering harmonious relationships among students within the class-group ( Marengo et al., 2018 ). Positive relationships with students, marked by low conflict and high closeness, are linked to students’ better adjustment to the school context ( Baker, 2006 ; Longobardi et al., 2019 ), greater academic commitment ( Longobardi et al., 2016 , 2019 ), and to prosocial behavior and reduced aggression ( Jungert et al., 2016 ; Marengo et al., 2018 ). Conversely, low-quality relationships with students can undermine the teachers’ ability to perform these functions.

Interestingly, the main effect of perceived teacher unfairness for proactive aggression was moderated by sex, indicating that this positive association was apparent only for the male group. This result is quite unexpected, especially in light of what emerged in Lenzi et al. (2014) study that did not find evidence of sex differences in the association between perceived teacher unfairness and a specific type of proactive aggressive behavior (i.e., bullying). This different finding may be explained by the different design and model of analysis of our study, by sample differences, and other potential methodological differences, but it is currently impossible to draw conclusions. Sex is indeed often identified as a significant moderator in the peer aggression literature, even though it is not always easy to understand what really makes some associations stronger for one sex group compared to the other. In sum, although exploratory, it remains an interesting result and it suggests that further research is certainly warranted to better explore how and under what circumstances perceiving to be treated unfairly at school may be differentially important for females’ and males’ adjustment and well-being.

4.2 Class-level effects

Using the lens of the socio-ecological model, another aim of this study was to explore the classroom as an important context that influences students’ behavior and satisfaction. Regarding average differences of peer aggression at the class-level, after controlling for the stability of the aggressive behavior, no significant predictor emerged. This did not fully confirm what was found in the few previous studies that have analyzed—cross-sectionally—the role of class teacher unfairness in students’ bullying and violence ( Vieno et al., 2011 ). The lack of statistically significant effects in this study may, of course, have different explanations. The first is the relative stability of the two class-level variables over the short period of time considered in our project. Moreover, even though there was a certain degree of between-class variability in reactive and, especially, proactive aggression, which justified class-level analysis, the majority of the variance of aggressive behavior was at the individual-level. This is a very common pattern in the school aggression literature, which may partly explain why it is not always easy to find significant predictors at the class-level. Finally, research on peer aggression at school has consistently identified the significant role of other class-level predictors, even within the realm of negative teacher-student relationships (e.g., Thornberg et al., 2018 ; Ten Bokkel et al., 2023 ). An additional, not necessarily alternative possibility for this lack of significant findings, therefore, may be that other class-level variables (e.g., class attitudes and norms) play a more central role compared to perceived relational teacher unfairness in differentiating class-level ratings of peer aggression, while teacher unfairness may demonstrate its negative impact more at the individual-level of analysis.

Regarding school satisfaction, instead, class-level analysis confirmed that there was a large degree of between-class variability in school satisfaction at T2 and that this variability was partially explained by class-level teacher unfairness. That is, not only individual students reported to be less satisfied if they felt to be treated unfairly by their teachers but, on average, school satisfaction toward the end of the school year was found to be significantly lower in classrooms where classmates shared a perception of unfair treatment. This effect was quite strong and robust against the control variables and is consistent with findings from studies with adult samples, mainly within organizations, showing that people satisfaction is associated not only with their individual perceptions of being treated fairly, but also with shared perceptions of a “fair context” (i.e., the average of individual perceptions of fairness within a group; e.g., Mossholder et al., 1998 ; Naumann and Bennett, 2000 ). This is an important novel contribution to the still limited literature on teacher relational unfairness within the classroom, showing for the first time a longitudinal negative effect of class-level teacher unfairness on students’ school satisfaction during one school year.

4.3 Limitations and implications

One limitation of this study is the relatively brief time span between the first and second waves. While the study made a significant contribution by examining both the individual-level and class-level impacts of perceived teacher unfairness over time, the conclusions drawn from our findings are limited to a period of approximately 6 months. Since we did not gather data over multiple years, we were unable to investigate how the relationships between the predictors and outcomes might evolve over a more extended period. It is recommended that future research expand beyond a single school year and potentially track adolescents over several years to gain a deeper understanding of these dynamics.

Second, all variables were evaluated using self-report questionnaires. It should be noted that self-report data may be influenced by social desirability bias, although it is often the most suitable or only option for researchers. In the context of this study, the primary focus was on the beliefs and perceptions of adolescents, which can only be conveyed by the individuals themselves. Moreover, relying solely on peer nominations or ratings to assess aggressive behavior may not always be reliable, as students may not accurately report their peers’ aggressive tendencies. In certain instances, individuals possess exclusive knowledge of specific aspects of their behavior, such as covert aggression. However, it is crucial to verify the findings of this study through longitudinal research that incorporates multiple sources of information regarding students’ aggressive behavior. Furthermore, including insights from teachers on their relationships with students, particularly in terms of fairness, could provide informative perspectives and enhance our understanding of classroom dynamics. Regarding teacher unfairness, a future line of research would be to compare students’ beliefs with ‘reality,’ to test whether different patterns of results emerge between students who believe that they are treated unfairly by their teacher, which is not true, and those who also have the same perception, but which corresponds to actual unfair treatment by the teacher.

Finally, in accordance with prior studies ( Engels et al., 2016 ), students were asked to provide feedback on their overall experience with teachers in the classroom, rather than focusing on a specific teacher. This decision was made due to the fact that students interact with multiple teachers on a daily basis, making it impractical to gather individual feedback for each teacher. While this approach offers a broad understanding of how adolescents perceive their relationship with teachers, it is important to acknowledge that it may lack specificity. Consequently, this limits the ability to examine if certain teacher characteristics (such as sex, years of experience, or teaching hours per week) influence the connection between teacher unfairness and adolescent school adjustment. Moreover, it restricts the exploration of different patterns of teacher relationships, such as having one teacher who is fair and supportive while others are not, or having all teachers who are fair and supportive.

Despite the limitations of this study, the findings suggest that minimizing teacher unfairness is important for reducing student aggression and improving their school satisfaction. Effective conflict reduction strategies can help teachers recognize students’ good behavior and achievements, set high expectations, and build constructive relationships with individual students ( Stipek and Miles, 2008 ). For example, teachers can negotiate behavioral contracts with students with whom they tend to experience the most conflict. These contracts should outline agreed-upon criteria for appropriate classroom behavior and interactions with the teacher and classmates ( Bowman-Perrott et al., 2015 ). Additionally, teachers can use “connective instruction,” particularly interpersonal connectiveness, as a strategy to build positive relationships with their students ( Martin and Dowson, 2009 ). The task at hand entails the active engagement of listening attentively to the perspectives expressed by students. It requires providing them with the opportunity to contribute to decisions that directly impact their lives. Moreover, it necessitates treating all students with equality, avoiding any form of bias, and affirming their worth. It also involves embracing and respecting their unique qualities and characteristics. Lastly, it entails maintaining a positive outlook and setting achievable goals for each student ( Flink et al., 1990 ; Teven and McCroskey, 1997 ; Slade, 2001 ). Considering the moderating role of students’ gender, at least for some variables, tailored strategies should also be implemented that take into account the possible differential role of teacher unfairness for different gender groups.

Finally, schools should implement activities able to promote and increase a sense of school belonging which can help to mitigate the negative impact of negative teacher-student relationships on students’ school satisfaction. One promising strategy is to use “small learning communities” ( Tillery et al., 2013 ). Briefly, these communities encompass the practice of organizing students into smaller networks within the broader school environment. These networks share common objectives and support systems, fostering a sense of connection and belonging among both students and adults. Consequently, educational institutions may contemplate implementing such initiatives to enhance the overall sense of community and affiliation within their schools.

Author’s note

The data used in this study were extracted from a larger dataset of a longitudinal project examining the social-cognitive factors associated with aggressive behavior in adolescents. A portion of this dataset was previously utilized in another study on the moral predictors of aggressive behavior ( Gini et al., 2022 ). Although the sample is the same, there is minimal overlap between the data used in the previous study and the current study, with only age, sex, and reactive and proactive aggression being common variables, serving different research purposes. The data on teacher unfairness, school satisfaction, and school stress have never been used in previously published manuscript. Moreover, the theoretical frameworks, the respective literatures, and the specific research questions of the two manuscripts are different.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by Ethics Committee for Psychological Research of the University of Padova. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants' legal guardians/next of kin.

Author contributions

GG: Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft. FA: Data curation, Writing – review & editing. TP: Conceptualization, Writing – original draft.

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. Open Access funding provided by Università degli Studi di Padova | University of Padua, Open Science Committee.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Adams, J. S. (1965). “Inequity in social exchange” in Advances in experimental social psychology . ed. L. Berkowitz (New York: Academic Press), 267–299.

Google Scholar

Arsenio, W., and Gold, J. (2006). The effects of social injustice and inequality on children’s moral judgments and behavior: towards a theoretical model. Cogn. Dev. 21, 388–400. doi: 10.1016/j.cogdev.2006.06.005

Crossref Full Text | Google Scholar

Baker, J. A. (2006). Contributions of teacher–child relationships to positive school adjustment during elementary school. J. Sch. Psychol. 44, 211–229. doi: 10.1016/j.jsp.2006.02.002

Bayram Özdemir, S., and Özdemir, M. (2020). How do adolescents’ perceptions of relationships with teachers change during upper-secondary school years? J. Youth Adolesc. 49, 921–935. doi: 10.1007/s10964-019-01155-3

PubMed Abstract | Crossref Full Text | Google Scholar

Bowman-Perrott, L., Burke, M. D., de Marin, S., Zhang, N., and Davis, H. (2015). A meta-analysis of single-case research on behavior contracts: effects on behavioral and academic outcomes among children and youth. Behav. Modif. 39, 247–269. doi: 10.1177/014544551455138

Bronfenbrenner, U. (1979). The ecology of human development: experiments by nature and design . Cambridge, MA: Harvard University Press.

Busching, R., and Krahé, B. (2020). With a little help from their peers: the impact of classmates on adolescents’ development of prosocial behavior. J. Youth Adolesc. 49, 1849–1863. doi: 10.1007/s10964-020-01260-8

Byrne, D. G., Davenport, S. C., and Mazanov, J. (2007). Profiles of adolescent stress: the development of the adolescent stress questionnaire (ASQ). J. Adolesc. 30, 393–416. doi: 10.1016/j.adolescence.2006.04.004

Card, N. A., Stucky, B. D., Sawalani, G. M., and Little, T. D. (2008). Direct and indirect aggression during childhood and adolescence: a meta-analytic review of gender differences, intercorrelations, and relations to maladjustment. Child Dev. 79, 1185–1229. doi: 10.1111/j.1467-8624.2008.01184.x

Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Struct. Equ. Model. Multidiscip. J. 14, 464–504. doi: 10.1080/10705510701301834

Cheung, G. W., and Rensvold, R. B. (2002). Evaluating goodness of fit indexes for testing measurement invariance. Struct. Equ. Model. A Multidisc. 9, 233–255. doi: 10.1207/S15328007SEM0902_5

Colquitt, J. A. (2001). On the dimensionality of organizational justice: a construct validation of a measure. J. Appl. Psychol. 86, 386–400. doi: 10.1037/0021-9010.86.3.386

Crick, N. R., and Dodge, K. A. (1994). A review and reformulation of social information-processing mechanisms in children's social adjustment. Psychol. Bull. 115, 74–101. doi: 10.1037/0033-2909.115.1.74

Cropanzano, R., Byrne, Z. S., Bobocel, D. R., and Rupp, D. E. (2001). Moral virtues, fairness heuristics, social entities, and other denizens of organizational justice. J. Vocat. Behav. 58, 164–209. doi: 10.1006/jvbe.2001.1791

Danielsen, A. G., Wiium, N., Wilhelmsen, B. U., and Wold, B. (2010). Perceived support provided teachers and classmates and students' self-reported academic initiative. J. Sch. Psychol. 48, 247–267. doi: 10.1016/j.jsp.2010.02.002

Enders, C. K., and Bandalos, D. L. (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Struct. Equ. Model. 8, 430–457. doi: 10.1207/S15328007SEM0803_5

Engels, M. C., Colpin, H., Van Leeuwen, K., Bijttebier, P., Van Den Noortgate, W., Claes, S., et al. (2016). Behavioral engagement, peer status, and teacher–student relationships in adolescence: a longitudinal study on reciprocal influences. J. Youth Adolesc. 45, 1192–1207. doi: 10.1007/s10964-016-0414-5

Flink, C., Boggiano, A. K., and Barrett, M. (1990). Controlling teaching strategies: undermining children’s self-determination and performance. J. Pers. Soc. Psychol. 59, 916–924. doi: 10.1037/0022-3514.59.5.916

Gilman, R., Huebner, E. S., and Laughlin, J. (2000). A first study of the multidimensional Students' life scale with adolescents. Soc. Indic. Res. 52, 135–160. doi: 10.1023/A:1007059227507

Gini, G., Marino, C., Pozzoli, T., and Holt, M. (2018). Associations between peer victimization, perceived teacher unfairness, and adolescents' adjustment and well-being. J. Sch. Psychol. 67, 56–68. doi: 10.1016/j.jsp.2017.09.005

Gini, G., Pozzoli, T., and Bussey, K. (2015). Moral disengagement moderates the link between psychopathic traits and aggressive behavior among early adolescents. Merrill-Palmer Q. 61, 51–67. doi: 10.13110/merrpalmquar1982.61.1.0051

Gini, G., Thornberg, R., Bussey, K., Angelini, F., and Pozzoli, T. (2022). Longitudinal links of individual and collective morality with adolescents’ peer aggression. J. Youth Adolesc. 51, 524–539. doi: 10.1007/s10964-021-01518-9

Hjern, A., Alfven, G., and Östberg, V. (2008). School stressors, psychological complaints and psychosomatic pain. Acta Paediatr. 97, 112–117. doi: 10.1111/j.1651-2227.2007.00585.x

Huebner, E. S. (1994). Preliminary development and validation of a multidimensional life satisfaction scale for children. Psychol. Assess. 6, 149–158. doi: 10.1037/1040-3590.6.2.149

James, K., Bunch, J., and Clay-Warner, J. (2015). Perceived injustice and school violence: an application of general strain theory. Youth Violence Juvenile Justice 13, 169–189. doi: 10.1177/1541204014521251

Jungert, T., Piroddi, B., and Thornberg, R. (2016). Early adolescents’ motivations to defend victims in school bullying and their perceptions of student–teacher relationships: a self-determination theory approach. J. Adolesc. 53, 75–90. doi: 10.1016/j.adolescence.2016.09.001

Kazemi, A. (2016). Examining the interplay of justice perceptions, motivation, and school achievement among secondary school students. Soc. Justice Res 29, 103–118. doi: 10.1007/s11211-016-0261-2

Krause, A., and Smith, J. D. (2022). Peer aggression and conflictual teacher-student relationships: a meta-analysis. Sch. Ment. Heal. 14, 306–327. doi: 10.1007/s12310-021-09483-1

Lazarus, R. S. (1966). Psychological stress and the coping process . New York: McGraw-Hill.

Lenzi, M., Vieno, A., De Vogli, R., Santinello, M., Ottova, V., Baška, T., et al. (2013). Perceived teacher unfairness and headache in adolescence: a cross-national comparison. Int. J. Public Health 58, 227–235. doi: 10.1007/s00038-012-0345-1

Lenzi, M., Vieno, A., Gini, G., Pozzoli, T., Pastore, M., Santinello, M., et al. (2014). Perceived teacher unfairness, instrumental goals, and bullying behavior in early adolescence. J. Interpers. Violence 29, 1834–1849. doi: 10.1177/0886260513511694

Little, T. D., Brauner, J., Jones, S. M., Nock, M. K., and Hawley, P. H. (2003). Rethinking aggression: a typological examination of the functions of aggression. Merrill-Palmer Q. 49, 343–369. doi: 10.1353/mpq.2003.0014

Löfstedt, P., García-Moya, I., Corell, M., Paniagua, C., Samdal, O., Välimaa, R., et al. (2020). School satisfaction and school pressure in the WHO European region and North America: an analysis of time trends (2002–2018) and patterns of co-occurrence in 32 countries. J. Adolesc. Health 66, S59–S69. doi: 10.1016/j.jadohealth.2020.03.007

Longobardi, C., Prino, L. E., Marengo, D., and Settanni, M. (2016). Student-teacher relationships as a protective factor for school adjustment during the transition from middle to high school. Front. Psychol. 7:1988. doi: 10.3389/fpsyg.2016.01988

Longobardi, C., Settanni, M., Prino, L. E., Fabris, M. A., and Marengo, D. (2019). Students’ psychological adjustment in normative school transitions from kindergarten to high school: investigating the role of teacher-student relationship quality. Front. Psychol. 10:1238. doi: 10.3389/fpsyg.2019.01238

Marengo, D., Jungert, T., Iotti, N. O., Settanni, M., Thornberg, R., and Longobardi, C. (2018). Conflictual student–teacher relationship, emotional and behavioral problems, prosocial behavior, and their associations with bullies, victims, and bullies/victims. Educ. Psychol. 38, 1201–1217. doi: 10.1080/01443410.2018.1481199

Martin, A. J., and Dowson, M. (2009). Interpersonal relationships, motivation, engagement, and achievement: yields for theory, current issues, and educational practice. Rev. Educ. Res. 79, 327–365. doi: 10.3102/00346543083255

MIUR (2019). MIUR – Ufficio di statistica: Gli alunni con cittadinanza non italiana (A.S. 2019/2020) . Rome, Italy: MIUR.

Mossholder, K. W., Bennett, N., and Martin, C. L. (1998). A multilevel analysis of procedural justice context. J. Organ. Behav. 19, 131–141. doi: 10.1002/(SICI)1099-1379(199803)19:2<131::AID-JOB878>3.0.CO;2-P

Murdock, T. B. (1999). The social context of risk: status and motivational predictors of alienation in middle school. J. Educ. Psychol. 91, 62–75. doi: 10.1037/0022-0663.91.1.62

Muthén, L. K., and Muthén, B. O. (1998–2017). Mplus user's guide (7th). Los Angeles, CA: Muthén & Muthén.

Naumann, S. E., and Bennett, N. (2000). A case for procedural justice climate: development and test of a multilevel model. Acad. Manag. J. 43, 881–889. doi: 10.5465/1556416

Osterman, K. F. (2000). Students' need for belonging in the school community. Rev. Educ. Res. 70, 323–367. doi: 10.3102/00346543070003323

Polman, H., De Castro, B. O., Koops, W., van Boxtel, H. W., and Merk, W. W. (2007). A meta-analysis of the distinction between reactive and proactive aggression in children and adolescents. J. Abnorm. Child Psychol. 35, 522–535. doi: 10.1007/s10802-007-9109-4

Rasooli, A., Zandi, H., and DeLuca, C. (2019). Conceptualising fairness in classroom assessment: exploring the value of organizational justice theory. Assess. Educ. Princ. Policy Pract. 26, 584–611. doi: 10.1080/0969594X.2019.1593105

Ren, P., Wang, Y., Liang, Y., Li, S., and Wang, Q. (2023). Bidirectional relationship between bullying victimization and functions of aggression in adolescents: the mediating effect of teacher justice. J. Adolesc. 95, 1245–1257. doi: 10.1002/jad.12198

Ripski, M. B., and Gregory, A. (2009). Unfair, unsafe, and unwelcome: do high school students' perceptions of unfairness, hostility, and victimization in school predict engagement and achievement? J. Sch. Violence 8, 355–375. doi: 10.1080/15388220903132755

Roeser, R. W., Eccles, J., and Sameroff, A. J. (1998). Academic and emotional functioning in early adolescence: longitudinal relations, patterns, and prediction by experience in middle school. Dev. Psychopathol. 10, 321–352. doi: 10.1017/S0954579498001631

Sabbagh, C., and Resh, N. (2016). Unfolding justice research in the realm of education. Soc. Justice Res 29, 1–13. doi: 10.1007/s11211-016-0262-1

Salmivalli, C., and Voeten, M. (2004). Connections between attitudes, group norms, and behaviour in bullying situations. Int. J. Behav. Dev. 28, 246–258. doi: 10.1080/01650250344000488

Samdal, O., Nutbeam, B., Wold, B., and Kannas, L. (1998). Achieving health and educational goals through schools—a study of the importance of school climate and the students' satisfaction with school. Health Educ. Res. Theory Pract. 13, 383–397. doi: 10.1093/her/13.3.383

Santinello, M., Vieno, A., and De Vogli, R. (2009). Primary headache in Italian early adolescents: the role of perceived teacher unfairness. Headache 49, 366–374. doi: 10.1111/j.1526-4610.2008.01208.x

Santinello, M., Vieno, A., and De Vogli, R. (2011). Bullying in Italian schools: the role of perceived teacher unfairness. Eur. J. Psychol. Educ. 26, 235–246. doi: 10.1007/s10212-010-0050-5

Skarlicki, D. P., and Folger, R. (1997). Retaliation in the workplace: the roles of distributive, procedural, and interactional justice. J. Appl. Psychol. 82, 434–443. doi: 10.1037/0021-9010.82.3.434

Slade, M. (2001). Listening to boys. Boys Sch. Bull. 4, 10–18.

Stipek, D., and Miles, S. (2008). Effects of aggression on achievement: does conflict with the teacher make it worse? Child Dev. 79, 1721–1735. doi: 10.1111/j.1467-8624.2008.01221.x

Stipek, D., Salmon, J. M., Givvin, K. B., Kazemi, E., Saxe, G., and MacGyvers, V. L. (1998). The value (and convergence) of practices suggested by motivation research and promoted by mathematics education reformers. J. Res. Math. Educ. 29, 465–488. doi: 10.2307/749862

Swearer, S. M., and Hymel, S. (2015). Understanding the psychology of bullying: moving toward a social-ecological diathesis–stress model. Am. Psychol. 70, 344–353. doi: 10.1037/a0038929

Szumski, G., Smogorzewska, J., and Grygiel, P. (2020). Attitudes of students toward people with disabilities, moral identity and inclusive education-a two-level analysis. Res. Dev. Disabil. 102:103685. doi: 10.1016/j.ridd.2020.103685

Ten Bokkel, I. M., Roorda, D. L., Maes, M., Verschueren, K., and Colpin, H. (2023). The role of affective teacher–student relationships in bullying and peer victimization: a multilevel meta-analysis. Sch. Psychol. Rev. 52, 110–129. doi: 10.1080/2372966X.2022.2029218

Teven, J. J., and McCroskey, J. C. (1997). The relationship of perceived teacher caring with student learning and teacher evaluation. Commun. Educ. 46, 1–9. doi: 10.1080/03634529709379069

Thornberg, T., Wänström, L., Pozzoli, T., and Gini, G. (2018). Victim prevalence in bullying and its association with teacher–student and student–student relationships and class moral disengagement: a class-level path analysis. Res. Pap. Educ. 33, 320–335. doi: 10.1080/02671522.2017.1302499

Tillery, A. D., Varjas, K., Roach, A. T., Kuperminc, G. P., and Meyers, J. (2013). The importance of adult connections in adolescents' sense of school belonging: implications for schools and practitioners. J. Sch. Violence 12, 134–155. doi: 10.1080/15388220.2012.762518

Torsheim, T., Cavallo, F., Levin, K. A., Schnohr, C., Mazur, J., Niclasen, B., et al. (2016). Psychometric validation of the revised family affluence scale: a latent variable approach. Child Indic. Res. 9, 771–784. doi: 10.1007/s12187-015-9339-x

Tyler, T. R., and Lind, E. A. (1992). A relational model of authority in groups. Adv. Exp. Soc. Psychol. 25, 115–192. doi: 10.1016/S0065-2601(08)60283-X

Tyler, T. R., and Smith, H. J. (1999). “Justice, social identity, and group processes” in The psychology of the social self . eds. T. R. Tyler, R. M. Kramer, and O. P. John (Mahwah, New Jersey: Lawrence Erlbaum Associates), 223–264.

Vieno, A., Gini, G., Santinello, M., Lenzi, M., and Nation, M. (2011). Violent behavior and unfairness in school: multilevel analysis of Italian schools. J. Community Psychol. 39, 534–550. doi: 10.1002/jcop.20450

Wendorf, C. A., and Alexander, S. (2005). The influence of individual- and class-level fairness-related perceptions on student satisfaction. Contemp. Educ. Psychol. 30, 190–206. doi: 10.1016/j.cedpsych.2004.07.003

Keywords: perceived teacher unfairness, reactive aggression, proactive aggression, school satisfaction, teacher injustice

Citation: Gini G, Angelini F and Pozzoli T (2024) Unfair teachers, unhappy students: longitudinal associations of perceived teacher relational unfairness with adolescent peer aggression and school satisfaction. Front. Psychol . 15:1321050. doi: 10.3389/fpsyg.2024.1321050

Received: 13 October 2023; Accepted: 04 April 2024; Published: 19 April 2024.

Reviewed by:

Copyright © 2024 Gini, Angelini and Pozzoli. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Gianluca Gini, [email protected]

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Jessica Grose

Most teachers know they’re playing with fire when they use tech in the classroom.

By Jessica Grose

Opinion Writer

A few years ago, when researchers at Boston College and Harvard set out to review all of the existing research on educational apps for kids in preschool through third grade, they were surprised to find that even though there are hundreds of thousands of apps out there that are categorized as educational, there were only 36 studies of educational apps in the databases they searched. “That is not a strong evidence base on which to completely redesign an entire schooling system,” Josh Gilbert, one of the co-authors of the study, told me over the phone.

That said, their meta-analysis of the effects of educational app use on children’s literacy and math skills, published in 2021, found that well-designed apps can make a positive difference when it comes to “constrained skills” — things like number recognition or times tables in math, or letter sounds in literacy. Unconstrained skills are more complex ones that develop over a lifetime of learning and can deepen over the years. (It’s worth noting that many popular educational apps are not high-quality .)

Gilbert said that overall, “the range of effects was gigantic.” Because they were all over the place, “we have to go beyond the average effect and say, OK, for whom does the app work? Under what conditions? On what types of measures? And I think those are the questions that researchers, policymakers, school leaders, teachers and principals should be asking,” he said. “What are the best use cases for this digital technology in the classroom?”

In last week’s newsletter , I came in pretty hot about the pitfalls of educational technology in American classrooms. I’m convinced that since students returned to in-person school after the disruptions of 2020-21, there are too many schools that haven’t been taking a thoughtful or evidence-based approach to how they’re using screens and apps, and that it’s time for a pause and a rethink. But that doesn’t mean there are no benefits to any use of educational technology.

So for the second part of this series, I wanted to talk to people who’ve seen real upsides from using tech in their classrooms. Their experiences back up some of the available research , which shows that ed tech can help teachers differentiate their material to meet the needs of students with a wide range of proficiencies. Further, teachers report that students with disabilities can really benefit from the assistive technologies that screens and apps can provide.

Debbie Marks, who teaches third grade in Oklahoma, told me that her students’ school-issued laptops allow them “to participate in differentiated reading interventions designed specifically for them” during the school day. That differentiation allows her to better assess how each student has progressed and tailor her instruction to each student.

“So for example, we could be working on story elements and we’re working on characters,” she explained to me when we spoke. “One student might be at the point where they’re just trying to identify who the main character is. Another student might be trying to identify character traits while a higher-level student would be comparing characters or would be identifying how the character changes throughout the story based on the plot. So it really allows me to develop one-on-one lessons for every kid in my classroom.”

Marks works in a rural district, about 90 minutes away from Tulsa, and some of her students may be traveling 45 minutes to an hour just to get to class. She said that the use of devices allows her to better connect with her students’ parents and to get them more involved in what’s going on in a classroom that is physically far from them. Marks also said that screens enable her to do things like virtual author visits, which she says get the kids really excited and engaged in reading.

I also heard from several teachers who said that assistive technology has been a game changer for students with special needs. Duncan Law, who works as a special education support teacher in an elementary school in Oregon, put it this way: “Technology can be a necessity for students with special needs in accessing core curriculum/standards, as well as for fluency practice. In the best case scenario, learning via tech is guided and closely monitored by teachers, and students are actively engaged with feedback. For students with dysgraphia and dyslexia, word processing tools offer a meaningful way to demonstrate/assess their writing skills.”

Several middle school and high school teachers who said that tech was helpful in their classrooms seemed to be using it as an efficient way to teach students more rote tasks, allowing more class time to be spent helping build those “unconstrained” skills.

Doug Showley, a high school English teacher in Indiana who’s been teaching since 1996, gave me the example of how he has changed his quizzes over time by integrating technology. He used to just give straight-up vocabulary quizzes where students had to define words; now he and his colleagues have moved toward “diction quizzes,” requiring students to understand the nuances of using specific words in sentences.

Showley noted that it’s easier to quickly look up words than it was in the hard-copy dictionary days, and that his students “have access to online dictionaries” during these quizzes. They’re given four synonyms and are asked to figure out which synonym best fits into a sentence. “To determine that, they have to go beyond just that basic definition. They’ve got to get into the connotative meaning of the word and the common usage of the word,” he explained.

But Showley also said that he monitors the kids quite closely. When they’re doing a task that involves their laptops, he’ll have them set up so all of their screens are facing him. He estimates that usually only one or two kids out of a class of 25 really aren’t able to stay on task when they’re on the screens.

He also told me that his school has made the decision not to block A.I., including ChatGPT, though it is a hot topic of discussion. The challenge of dealing with A.I. is something that came up a lot among teachers in the upper grades, and the overall vibe I got was that no one quite knows what to do with it yet.

After we spoke, Showley emailed me to say that “we should carefully gauge to what degree and in what way tech is used at each level of education.” And he wrote something that I think really sums up both the promise and the peril of ed tech (and is also such a classic English teacher passage):

I couldn’t help but think of Prometheus defying the Olympic gods by sharing the first-ever technological advancement with humankind: fire. Fire, as with every other significant advancement since, both propelled society forward and burnt it to the ground. It enlightened our minds and souls, and it tormented them, just as Prometheus was perpetually tormented through his punishment for sharing too much of the gods’ power.

Perhaps deliberately, one of the popular digital whiteboards is the Promethean board.

The technology isn’t going away. We need to start creating better frameworks to think about how students and teachers are using technology in our schools, because the tech companies won’t stop pushing their products, whether or not there’s evidence that shows educational gains. CNN’s Clare Duffy reports that later this year, Meta “will launch new software for educators that aims to make it easier to use its V.R. headsets in the classroom,” though “it remains unclear just how useful virtual reality is in helping students learn better.”

In next week’s newsletter, I’ll write about solutions to some of the problems posed by ed tech, and how we might create a future where we can minimize some of the most egregious hazards of distraction and invasion of privacy, and realize some of the potential of technology’s most fantastic educational promises.

Jessica Grose is an Opinion writer for The Times, covering family, religion, education, culture and the way we live now.

Research news: Why are young people leaving school early?

If you’re a secondary teacher or leader, what is the average year 12 completion rate for students in your school? Do you know the reasoning behind why some students decide to leave early? How could you best support these students? A new report offers answers to these questions.

The Pathways, Engagement and Transitions (PET) study by The Smith Family examines the post-school pathways of young people experiencing disadvantage. The study’s latest report, Experiences of early school leavers , has just been released and is the third in the PET series. It focuses on year 12 completion of students who were in year 10 in 2020. Overall, there were 2,000 participants involved who completed surveys in 2021, 2022 and 2023. In-depth interviews were conducted with 29 of the participants.

Of these participants, 68% completed year 12, 28% left school early and 4% were still in school when they completed the 2023 survey. A majority of the early school leavers surveyed (65%) had completed year 11 and a smaller number (35%) had completed year 10. The report also shares how things are looking nationally, reporting that year 12 completion rates declined from 79% in 2021 to 76% in 2022.

Crucially, the report found that 92% of students who didn’t complete year 12 said they intended to finish school when they were in year 11. ‘This means there’s a tremendous opportunity to help more young people to realise that goal, through better use of data and more individualised assistance including for literacy and numeracy, better support with mental health and quality careers support,’ Head of Research and Advocacy at The Smith Family, Anne Hampshire, says.

Their research also shows early school leavers were more likely to be male, Aboriginal and/or Torres Strait Islander, from an English-speaking background, to have a health and/or mental health condition and/or be living in a regional area.

Other key findings identify important indicators of year 12 completion, such as attendance, academic achievement and receiving careers advice.

Why young people are leaving school early

The survey findings reveal 8 factors identified by young people as contributing to why they left school early.

[The 8 reasons identified by young people for leaving school early: adapted from The Smith Family, 2024)

Half of all early school leavers identified multiple reasons for leaving school. ‘Those who offered only one reason were most likely to cite having got or wanted to get a job, apprenticeship or traineeship (22% and 17%, respectively) or having health or mental health issues (15%),’ the report reads.

‘Providing students experiencing these challenges with more individualised support while at school [can] strengthen school engagement and completion. The cohort of early school leavers is diverse and leave for a range of reasons, and require tailored approaches to address their needs.’

Careers and pathways education plays a crucial role

International research has shown the positive impacts careers education can have on student engagement and academic motivation. The PET study has found a correlation between a student’s experience of receiving careers advice and whether or not they complete year 12.

[School completion by experience of careers advice at school: adapted from The Smith Family, 2024]

In light of these findings, The Smith Family say increased provision of individualised career advice and support throughout secondary schooling years is needed. ‘Support should be accessible, engaging, and meaningful to young people. Tailored support can both contribute to greater levels of school engagement and completion and stronger post-school pathways for those young people who do leave school early,’ they write.

The report also includes quotes from young people on the topic of careers and pathways education:

We had meetings with the careers counsellor at our school and they would ask us what course we wanted to go into or what job we wanted to do in the future, and advise us on what uni to go to and what ATAR would be required. They also showed me the different pathways in case I didn’t get a high enough ATAR. I could have gotten a Diploma, Advanced Diploma, and so on. At first I had doubts I would get into the course I wanted, but that conversation helped set my mind at ease. It was very helpful so I’m grateful for that. – Ali* (Year 12 completer)

I think it could be more useful to explore different avenues with young people; spend more time talking about the future and what they actually want to do. I think that would definitely help. – Liam* (early school leaver)

*Names have been changed

Early indicators of students who may need extra support

A major finding from the report is the impact poor attendance has on a student’s likelihood to leave school early. Researchers found that over half of students whose attendance was considered low in year 9 left school before finishing year 12, compared to less than 1 in 5 students who had high attendance.

[School completion by year 9 attendance rate: adapted from The Smith Family, 2024]

As well as this, academic achievement appears to play a role in year 12 completion. The data show 45% of participants who received a D or E grade in year 9 English left school early, compared to 21% of students who received an A, B or C grade.

‘These findings reinforce that attendance and achievement can act as an early flag to identify young people needing additional support to complete year 12,’ the report says, adding that continuous monitoring of warning signs throughout school to identify young people at elevated risk of early school leaving is needed.

The PET study has also published reports on the dynamic post-school pathways of young people experiencing disadvantage and the initial post-school transitions of young people experiencing disadvantage .

Teacher has covered the ongoing PET study in a podcast episode featuring Anne Hampshire (Earp, 2022) and an article (Vukovic, 2023) focusing on post-school transitions of students.

Earp, J. (2022, November 2). The Research Files Episode 78: Supporting disadvantaged students in post-school pathways and transitions. Teacher magazine. https://www.teachermagazine.com/au_en/articles/the-research-files-episode-78-supporting-disadvantaged-students-in-post-school-pathways-and-transitions

The Smith Family. (2023). Pathways, engagement and transitions: Experiences of early school leavers.

Vukovic, R. (2023, August 11). Research update: Supporting post-school transition for disadvantaged students. Teacher magazine. https://www.teachermagazine.com/au_en/articles/research-update-supporting-post-school-transition-for-disadvantaged-students

The Smith Family says rates of attendance can be an early flag for identifying young people needing additional support to complete year 12. How often do you collect and review attendance data in your own school? How are you using data collected to inform the supports you provide to students?

Related articles

Numbers, Facts and Trends Shaping Your World

Read our research on:

Full Topic List

Regions & Countries

• Publications
• Our Methods
• Short Reads
• Tools & Resources

Read Our Research On:

Partisan divides over K-12 education in 8 charts

K-12 education is shaping up to be a key issue in the 2024 election cycle. Several prominent Republican leaders, including GOP presidential candidates, have sought to limit discussion of gender identity and race in schools , while the Biden administration has called for expanded protections for transgender students . The coronavirus pandemic also brought out partisan divides on many issues related to K-12 schools .

Today, the public is sharply divided along partisan lines on topics ranging from what should be taught in schools to how much influence parents should have over the curriculum. Here are eight charts that highlight partisan differences over K-12 education, based on recent surveys by Pew Research Center and external data.

Pew Research Center conducted this analysis to provide a snapshot of partisan divides in K-12 education in the run-up to the 2024 election. The analysis is based on data from various Center surveys and analyses conducted from 2021 to 2023, as well as survey data from Education Next, a research journal about education policy. Links to the methodology and questions for each survey or analysis can be found in the text of this analysis.

Most Democrats say K-12 schools are having a positive effect on the country , but a majority of Republicans say schools are having a negative effect, according to a Pew Research Center survey from October 2022. About seven-in-ten Democrats and Democratic-leaning independents (72%) said K-12 public schools were having a positive effect on the way things were going in the United States. About six-in-ten Republicans and GOP leaners (61%) said K-12 schools were having a negative effect.

About six-in-ten Democrats (62%) have a favorable opinion of the U.S. Department of Education , while a similar share of Republicans (65%) see it negatively, according to a March 2023 survey by the Center. Democrats and Republicans were more divided over the Department of Education than most of the other 15 federal departments and agencies the Center asked about.

In May 2023, after the survey was conducted, Republican lawmakers scrutinized the Department of Education’s priorities during a House Committee on Education and the Workforce hearing. The lawmakers pressed U.S. Secretary of Education Miguel Cardona on topics including transgender students’ participation in sports and how race-related concepts are taught in schools, while Democratic lawmakers focused on school shootings.

Partisan opinions of K-12 principals have become more divided. In a December 2021 Center survey, about three-quarters of Democrats (76%) expressed a great deal or fair amount of confidence in K-12 principals to act in the best interests of the public. A much smaller share of Republicans (52%) said the same. And nearly half of Republicans (47%) had not too much or no confidence at all in principals, compared with about a quarter of Democrats (24%).

This divide grew between April 2020 and December 2021. While confidence in K-12 principals declined significantly among people in both parties during that span, it fell by 27 percentage points among Republicans, compared with an 11-point decline among Democrats.

Democrats are much more likely than Republicans to say teachers’ unions are having a positive effect on schools. In a May 2022 survey by Education Next , 60% of Democrats said this, compared with 22% of Republicans. Meanwhile, 53% of Republicans and 17% of Democrats said that teachers’ unions were having a negative effect on schools. (In this survey, too, Democrats and Republicans include independents who lean toward each party.)

The 38-point difference between Democrats and Republicans on this question was the widest since Education Next first asked it in 2013. However, the gap has exceeded 30 points in four of the last five years for which data is available.

Republican and Democratic parents differ over how much influence they think governments, school boards and others should have on what K-12 schools teach. About half of Republican parents of K-12 students (52%) said in a fall 2022 Center survey that the federal government has too much influence on what their local public schools are teaching, compared with two-in-ten Democratic parents. Republican K-12 parents were also significantly more likely than their Democratic counterparts to say their state government (41% vs. 28%) and their local school board (30% vs. 17%) have too much influence.

On the other hand, more than four-in-ten Republican parents (44%) said parents themselves don’t have enough influence on what their local K-12 schools teach, compared with roughly a quarter of Democratic parents (23%). A larger share of Democratic parents – about a third (35%) – said teachers don’t have enough influence on what their local schools teach, compared with a quarter of Republican parents who held this view.

Republican and Democratic parents don’t agree on what their children should learn in school about certain topics. Take slavery, for example: While about nine-in-ten parents of K-12 students overall agreed in the fall 2022 survey that their children should learn about it in school, they differed by party over the specifics. About two-thirds of Republican K-12 parents said they would prefer that their children learn that slavery is part of American history but does not affect the position of Black people in American society today. On the other hand, 70% of Democratic parents said they would prefer for their children to learn that the legacy of slavery still affects the position of Black people in American society today.

Parents are also divided along partisan lines on the topics of gender identity, sex education and America’s position relative to other countries. Notably, 46% of Republican K-12 parents said their children should not learn about gender identity at all in school, compared with 28% of Democratic parents. Those shares were much larger than the shares of Republican and Democratic parents who said that their children should not learn about the other two topics in school.

Many Republican parents see a place for religion in public schools , whereas a majority of Democratic parents do not. About six-in-ten Republican parents of K-12 students (59%) said in the same survey that public school teachers should be allowed to lead students in Christian prayers, including 29% who said this should be the case even if prayers from other religions are not offered. In contrast, 63% of Democratic parents said that public school teachers should not be allowed to lead students in any type of prayers.

In June 2022, before the Center conducted the survey, the Supreme Court ruled in favor of a football coach at a public high school who had prayed with players at midfield after games. More recently, Texas lawmakers introduced several bills in the 2023 legislative session that would expand the role of religion in K-12 public schools in the state. Those proposals included a bill that would require the Ten Commandments to be displayed in every classroom, a bill that would allow schools to replace guidance counselors with chaplains, and a bill that would allow districts to mandate time during the school day for staff and students to pray and study religious materials.

Mentions of diversity, social-emotional learning and related topics in school mission statements are more common in Democratic areas than in Republican areas. K-12 mission statements from public schools in areas where the majority of residents voted Democratic in the 2020 general election are at least twice as likely as those in Republican-voting areas to include the words “diversity,” “equity” or “inclusion,” according to an April 2023 Pew Research Center analysis .

Also, about a third of mission statements in Democratic-voting areas (34%) use the word “social,” compared with a quarter of those in Republican-voting areas, and a similar gap exists for the word “emotional.” Like diversity, equity and inclusion, social-emotional learning is a contentious issue between Democrats and Republicans, even though most K-12 parents think it’s important for their children’s schools to teach these skills . Supporters argue that social-emotional learning helps address mental health needs and student well-being, but some critics consider it emotional manipulation and want it banned.

In contrast, there are broad similarities in school mission statements outside of these hot-button topics. Similar shares of mission statements in Democratic and Republican areas mention students’ future readiness, parent and community involvement, and providing a safe and healthy educational environment for students.

• Education & Politics
• Partisanship & Issues
• Politics & Policy

Jenn Hatfield is a writer/editor at Pew Research Center

Most Americans think U.S. K-12 STEM education isn’t above average, but test results paint a mixed picture

About 1 in 4 u.s. teachers say their school went into a gun-related lockdown in the last school year, about half of americans say public k-12 education is going in the wrong direction, what public k-12 teachers want americans to know about teaching, what’s it like to be a teacher in america today, most popular.

1615 L St. NW, Suite 800 Washington, DC 20036 USA (+1) 202-419-4300 | Main (+1) 202-857-8562 | Fax (+1) 202-419-4372 |  Media Inquiries

Research Topics

• Age & Generations
• Coronavirus (COVID-19)
• Economy & Work
• Family & Relationships
• Gender & LGBTQ
• Immigration & Migration
• International Affairs
• Internet & Technology
• Methodological Research
• News Habits & Media
• Non-U.S. Governments
• Other Topics
• Race & Ethnicity
• Email Newsletters

ABOUT PEW RESEARCH CENTER  Pew Research Center is a nonpartisan fact tank that informs the public about the issues, attitudes and trends shaping the world. It conducts public opinion polling, demographic research, media content analysis and other empirical social science research. Pew Research Center does not take policy positions. It is a subsidiary of  The Pew Charitable Trusts .

Copyright 2024 Pew Research Center

Terms & Conditions

Privacy Policy

Cookie Settings

Reprints, Permissions & Use Policy

Jump to main content

Explore the latest news from the university of Northern Colorado.

Research on Culturally Appropriate Music Plays Out In Middle School Band Class

• Arts-and-Culture

Music Education student researching culturally responsive pedagogy from a student-led perspective.

April 24, 2024 | By Brenda Gillen

Victoria DiMarzio teaches band and orchestra at a Denver middle school. Their students hail from 30 countries, including recent immigrants and refugees. In DiMarzio's inclusive and welcoming class, students weigh in on music that appropriately represents different cultures. DiMarzio is a Master of Music student in the University of Northern Colorado's Music Education Concentration program , where they are researching culturally responsive pedagogy from a student-led perspective.   

"I wanted to focus on culturally responsive education, and several UNC professors have the same kind of drive and focus on that topic." — Victoria DiMarzio

"The newcomer program at my school is designed for students who have interrupted education, so they spend most of their day focusing intensely on learning English. I've seen the breadth of culture and what it is like for them to try and navigate the world. I'm doing my part to make sure they're getting the best education they can," they said.  

DiMarzio has been teaching in Denver for about a decade. UNC's Extended Campus offers a hybrid in-person and online program. It allows her to take one course per semester in the fall and spring while continuing to teach and interact with fellow students and professors in person in the summertime.  

"I really liked the program design. I wanted to focus on culturally responsive education, and several UNC professors have the same kind of drive and focus on that topic," DiMarzio said.  

In their thesis, DiMarzio is designing a project-based learning unit where students engage in practice and reflection learning cycles. First, students examine skill-level-appropriate music labeled "multicultural." Then, they analyze its merits by asking if the music accurately represents the culture intended, if it stereotypes musical styles or creates biased sounds and if it’s more appropriate to modify music to fit an ensemble or experience music in other ways to be more respectful.  

"Culturally responsive education is an approach to teaching that requires self-reflection on the teacher's part about personal background and biases. And then understanding your students' backgrounds and cultures, to honor them, respect them and use them as leverage to help them learn better. What I want to do is involve my students in that conversation," they said.  

DiMarzio's students research music from around the world. They focus on a region to learn about its styles, traditions and instrumentation, then compare music meant to represent that area.  

"A student who studied traditional Chinese opera looked up music written for band class to represent Chinese music. They had to make decisions like whether this is accurately portrayed or more of a stereotype and whether we should perform it," they said. "Hopefully, the students all learn about something new and get a new respect for the people in our class and how different we all are."   

DiMarzio finds the ability to put into practice the lessons from their research appealing, as they can make changes in the classroom right away. They're working with students to create a flowchart for the decision-making process to share with other music teachers planning multicultural music. Hopefully, DiMarzio said, it will improve performances so the average person can experience more robust and culturally validating music.  

In the College of Performing and Visual Arts' School of Music , Visiting Assistant Professor Krissie Weimer mentored DiMarzio through the research process.   

"Victoria is self-motivated, knows what they're interested in working on and goes after it. They're also very thoughtful and reflective in their work. It's clear from their work and research that they are a very student-centered teacher who is interested in connecting with students on multiple levels," Weimer said.  

They noted that DiMarzio's bringing the conversations about appropriate literature, or repertoire-like pieces, to their students in an age-appropriate way is a unique take on culturally responsive pedagogy.   

"They came in clear on what they wanted to do from the first research class, and we took that topic and turned it into a thesis. It feels more like a collaboration with Victoria rather than where I have to move them along," she said.  

After graduating in May 2024, DiMarzio will take a break from studies but has yet to decide whether to pursue a doctorate.  

"I really love being a middle school music teacher. I find a lot of joy and career fulfillment. But if I want to see change happen in our field, it's probably better done helping the next generation of teachers," they said.  

More Stories

Acclaimed Artist and Performer Represents Spirit of the Times in Race-Related Dissertation

Este artículo no está en español..

Inspiring Inclusivity in Music Through Achieving Milestones and Pushing Boundaries

Consiguiendo importantes logros y eliminando barreras para inspirar a la inclusión a través de la música.

PHOTOS: Lunar New Year

PHOTOS: Taste of Africa

Follow unc social.

Contact UNC

Social media.

• UNC Overview
• Awards & Accolades
• Organizational Chart
• Strategic Plan
• Accreditation
• Student Consumer Information
• Sustainability
• UNC Newsroom
• Events Calendar
• UNC Magazine
• Submit News
• Submit an Event
• Marketing and Communications
• University Advancement

Page Last Updated: Today | Contact for this Page: Deanna Herbert

Privacy Policy | Affirmative Action/Equal Employment Opportunity/Title IX Policy & Coordinator

Suggestions or feedback?

MIT News | Massachusetts Institute of Technology

• Machine learning
• Social justice
• Black holes
• Classes and programs

Departments

• Aeronautics and Astronautics
• Brain and Cognitive Sciences
• Architecture
• Political Science
• Mechanical Engineering

Centers, Labs, & Programs

• Abdul Latif Jameel Poverty Action Lab (J-PAL)
• Picower Institute for Learning and Memory
• Lincoln Laboratory
• School of Architecture + Planning
• School of Engineering
• School of Humanities, Arts, and Social Sciences
• Sloan School of Management
• School of Science
• MIT Schwarzman College of Computing

Professor Emeritus Bernhardt Wuensch, crystallographer and esteemed educator, dies at 90

Press contact :.

Previous image Next image

MIT Professor Emeritus Bernhardt Wuensch ’55, SM ’57, PhD ’63, a crystallographer and beloved teacher whose warmth and dedication to ensuring his students mastered the complexities of a precise science matched the analytical rigor he applied to the study of crystals, died this month in Concord, Massachusetts. He was 90.

Remembered fondly for his fastidious attention to detail and his office stuffed with potted orchids and towers of papers, Wuensch was an expert in X-ray crystallography, which involves shooting X-ray beams at crystalline materials to determine their underlying structure. He did pioneering work in solid-state ionics, investigating the movement of charged particles in solids that underpins technologies critical for batteries, fuel cells, and sensors. In education, he carried out a major overhaul of the curriculum in what is today MIT’s Department of Materials Science and Engineering (DMSE).

Despite his wide-ranging research and teaching interests, colleagues and students said, he was a perfectionist who favored quality over quantity.

“All the work he did, he wasn’t in a hurry to get a lot of stuff done,” says DMSE’s Professor Harry Tuller. “But what he did, he wanted to ensure was correct and proper, and that was characteristic of his research.”

Born in Paterson, New Jersey, in 1933, Wuensch first arrived at MIT as a first-year undergraduate in the 1950s. He earned bachelor’s and master’s degrees in physics before switching to crystallography and earning a PhD from what was then the Department of Geology (now Earth, Atmospheric and Planetary Sciences). He joined the faculty of the Department of Metallurgy in 1964 and saw its name change twice over his 46 years, retiring from DMSE in 2011.

As a professor of ceramics, Wuensch was a part of the 20th-century shift from a traditional focus on metals and mining to a broader class of materials that included polymers, ceramics, semiconductors, and biomaterials. In a 1973 letter supporting his promotion to full professor, then-department head Walter Owen credits Wuensch for contributing to “a completely new approach to the teaching of the structure of materials.”

His research led to major advancements in understanding how atomic-level structures affect magnetic and electrical properties of materials. For example, Tuller says, he was one of the first to detail how the arrangement of atoms in fast-ion conductors — materials used in batteries, fuel cells, and other devices — influences their ability to swiftly conduct ions.

Wuensch was a leading light in other areas, including diffusion, the movement of ions in materials such as liquids or gases, and neutron diffraction, aiming neutrons at materials to collect information about their atomic and magnetic structure.

Tuller, a DMSE faculty member for 49 years, tapped Wuensch’s expertise to study zinc oxide, a material used to make varistors, semiconducting components that protect circuits from high-voltage surges of electricity. Together, Tuller and Wuensch found that in such materials ions move much more rapidly along the grain boundaries — the interfaces between the crystallites that make up these polycrystalline ceramic materials.

“It’s what happens at those grain boundaries that actually limits the power that would go through your computer during a voltage surge by instead short-circuiting the current through these devices,” Tuller says. He credited the partnership with Wuensch for the knowledge. “He was instrumental in helping us confirm that we could engineer those grain boundaries by taking advantage of the very rapid diffusivity of impurity elements along those boundaries.”

In recognition of his accomplishments, Wuensch was elected a fellow of the American Ceramics Society and the Mineralogical Society of America and belonged to other professional associations, including The Electrochemical Society and Materials Research Society. In 2003 he was awarded an honorary doctorate from South Korea’s Hanyang University for his work in crystallography and diffusion-related phenomena in ceramic materials.

“A great, great teacher”

Known as “Bernie” to friends and colleagues, Wuensch was equally at home in the laboratory and the classroom. “He instilled in several generations of young scientists this ability to think deeply, be very careful about their research, and be able to stand behind it,” Tuller says.

One of those scientists is Sossina Haile ’86, PhD ’92, the Walter P. Murphy Professor of Materials Science and Engineering at Northwestern University, a researcher of solid-state ionic materials who develops new types of fuel cells, devices that convert fuel into electricity.

Her introduction to Wuensch, in the 1980s, was his class 3.13 (Symmetry Theory). Haile was at first puzzled by the subject, the study of the symmetrical properties of crystals and their effects on material properties. The arrangements of atoms and molecules in a material is crucial for predicting how materials behave in different situations — whether they will be strong enough for certain uses, for example, or can conduct electricity — but to an undergraduate it was “a little esoteric.”

“I certainly remember thinking to myself, ‘What is this good for?’” Haile says with a laugh. She would later return to MIT as a PhD student working alongside Wuensch in his laboratory with a renewed perspective.

Previous item Next item

“He just made seemingly esoteric topics really interesting and was very astute in knowing whether or not a student understood.” Haile describes Wuensch’s articulate speech, “immaculate” handwriting, and detailed drawings of three-dimensional objects on the chalkboard. Haile notes that his sketches were so skillful that students felt disappointed when they looked at a figure they tried to copy in their notebooks.

“They couldn’t tell what it was,” Haile says. “It felt really clear during lecture, and it wasn’t clear afterwards because no one had a drawing as good as his.”

Carl Thompson, the Stavros V. Salapatas Professor in Materials Science and Engineering at DMSE, was another student of Wuensch’s who came away with a broadened outlook. In 3.13, Thompson recalls Wuensch asking students to look for symmetry outside of class, patterns in a brick wall or in subway station tiles. “He said, ‘This course will change the way you see the world,’ and it did. He was a great, great teacher.”

In a 2005 videorecorded session of 3.60 (Symmetry, Structure, and Tensor Properties of Materials), a graduate class that he taught for three decades, Wuensch writes his name on the board along with his telephone extension number, 6889, pointing out its rotational symmetry.

“You can pick it up, turn it head-over-heels by 180 degrees, and it’s mapped into coincidence with itself,” Wuensch said. “You might think I would have had to have fought for years to get it, an extension number like that, but no. It just happened to come my way.”

(The class can be watched in its entirety on MIT OpenCourseWare .)

Wuensch also had a whimsical sense of humor, which he often exercised in the margins of his students’ papers, Haile says. In a LinkedIn tribute to him, she recalled a time she sent him a research manuscript with figures that was missing Figure 5 but referred to it in the text, writing that it plotted conductivity versus temperature.

“Bernie noted that figures don’t plot; people do, and evidently Figure 5 was missing because ‘it was off plotting somewhere,’” Haile wrote.

Reflecting on Wuensch’s legacy in materials science and engineering, Haile says his knowledge of crystallography and the manual analysis and interpretation he did in his time was critical. Today, materials science students use crystallographic software that automates the algorithms and calculations.

“The current students don’t know that analysis but benefit from it because people like Bernie made sure it got into the common vernacular at the time when code was being put together,” Haile said.

A multifaceted tenure

Wuensch served DMSE and MIT in innumerable other ways, serving on departmental committees on curriculum development, graduate students, and policy, and on School of Engineering and Institute-level committees on education and foreign scholarships, among others. “He was always involved in any committee work he was asked to do,” Thompson says.

He was acting department head for six months starting in 1980, and in 1988-93 he was the director of the Center for Materials Science and Engineering, an earlier iteration of today’s Materials Research Center.

For all his contributions, there are few things Wuensch was better known for at MIT than his office in Building 13, which had shelves lined with multicolored crystal lattice models, representing the arrangements of atoms in materials, and orchids he took meticulous care of. And then there was the cityscape of papers, piled in heaps on the floor, on his desk, on pullout extensions. Thompson says walking into his office was like navigating a canyon.

“He had so many stacks of paper that he had no place to actually work at his desk, so he would put things on his lap — he would start writing on his lap,” Haile says. “I remember calling him at one point in time and talking to him, and I said, ‘Bernie, you’re writing this down on your lap, aren’t you?’ And he said, ‘In fact, yes, I am.’”

Wuensch was also known for his kindness and decency. Angelita Mireles, graduate academic administrator at DMSE, says he was a popular pick for graduate students assembling committees for their thesis area examinations, which test how prepared students are to conduct doctoral research, “because he was so nice.”

That said, he had exacting standards. “He expected near perfection from his students, and that made them a lot deeper,” Tuller says.

Outside of MIT, Wuensch enjoyed tending his garden; collecting minerals, gemstones, and rare coins; and reading spy novels. Other pastimes included fishing and clamming in Maine, splitting his own firewood, and traveling with his wife, Mary Jane.

Wuensch is survived by his wife; son Stefan Wuensch and wife Wendy Joseph; daughter Katrina Wuensch and partner Jason Staly; and grandchildren Noemi and Jack.

Friends and family are invited to a memorial service Sunday, April 28, at 1:30 p.m. at Duvall Chapel at 80 Deaconess Road in Concord, Massachusetts. Memories or condolences can be posted at obits.concordfuneral.com/bernhardt-wuensch .

Share this news article on:

Related links.

• Materials Research Laboratory
• Department of Materials Science and Engineering

Related Topics

• Education, teaching, academics
• Energy storage
• Materials science and engineering

Related Articles

A green hydrogen innovation for clean energy

Harry Tuller honored for career advancing solid-state chemistry and electrochemistry

Harry Tuller wins Egleston Medal for his electroceramics work

More mit news.

Study demonstrates efficacy of MIT-led Brave Behind Bars program

Read full story →

Bringing an investigator’s eye to complex social challenges

MIT announces 2024 Bose Grants

Circadian rhythms can influence drugs’ effectiveness

Ian Waitz named vice president for research

A closed-loop drug-delivery system could improve chemotherapy

• More news on MIT News homepage →

Massachusetts Institute of Technology 77 Massachusetts Avenue, Cambridge, MA, USA

• Map (opens in new window)
• Events (opens in new window)
• People (opens in new window)
• Careers (opens in new window)
• Accessibility
• Social Media Hub
• MIT on Facebook
• MIT on YouTube
• MIT on Instagram

• International edition
• Australia edition
• Europe edition

Louisiana’s flagship university lets oil firms influence research – for a price

Louisiana State University allowed Shell to influence studies after a $25m donation and sought funds from other fossil fuel firms

• This story is co-published with the Lens, a non-profit newsroom in New Orleans

For $5m, Louisiana’s flagship university will let an oil company weigh in on faculty research activities. Or, for $100,000, a corporation can participate in a research study, with “robust” reviewing powers and access to all resulting intellectual property.

Those are the conditions outlined in a boilerplate document that Louisiana State University’s fundraising arm circulated to oil majors and chemical companies affiliated with the Louisiana Chemical Association, an industry lobbying group, according to emails disclosed in response to a public records request by the Lens .

Records show that after Shell donated $25m in 2022 to LSU to create the Institute for Energy Innovation, the university gave the fossil fuel corporation license to influence research and coursework for the university’s new concentration in carbon capture, use and storage.

Afterward, LSU’s fundraising entity, the LSU Foundation, used this partnership as a model to shop around to members of the Louisiana Chemical Association, such as ExxonMobil , Air Products and CF Industries, which have proposed carbon capture projects in Louisiana.

For $2m, Exxon became the institute’s first “strategic partner-level donor”, a position that came with robust review of academic study output and with the ability to focus research activities. Another eight companies have discussed similar deals with LSU, according to a partnership update that LSU sent to Shell last summer.

Some students, academics and experts said such relationships raise questions about academic freedom and public trust.

Asked to comment, the Institute for Energy Innovation’s director, Brad Ives, defended the partnerships, as did the oil majors. Two more companies have since entered into partnerships with the Institute for Energy Innovation, said Ives. But Shell is the only company to have donated at the level that gave the company a seat on the advisory board that chooses the institute’s research. The head of the Louisiana Chemical Association and the Louisiana Mid-Continent Oil and Gas Association also sit on the advisory board, which can vote to stop a research project from moving forward.

Ives said being able to work with oil and gas companies is “really a key to advancing energy innovation”.

A spokesperson for Shell said: “We’re proud to partner with LSU to contribute to the growing compendium of peer-reviewed climate science and advance the effort to identify multiple pathways that can lead to more energy with fewer emissions.”

An ExxonMobil spokesperson said: “Our collaboration with LSU and the Institute for Energy Innovation includes an allocation for research in carbon capture utilization and storage, as well as advanced recycling studies.”

LSU has long had a close-relationship with oil majors, the names of which hang from buildings and equipment at the university. Nearly 40% of LSU funding comes from the state, which received a good chunk of its revenue from oil and gas activities until the 1980s. In recent years, oil and gas revenue has made up less than 10% of the state budget.

But the new, highly visible partnership with Shell took the closeness a step further, promising corporations voting power over the Institute for Energy Innovation’s research activities in return for their investment.

“I have a hard time seeing a faculty member engaged in legitimate research being eager for an oil company or representative of a chemical company to vote on his or her research agenda,” said Robert Mann, political commentator and former LSU journalism professor . “That is an egregious violation of academic freedom.

“You don’t expect to see it written down like that,” Mann said, after the Lens asked him to review the boilerplate document that outlines what companies can expect in return for their donations to LSU’s Institute for Energy Innovation. It is not appropriate, Mann said, for faculty research to be driven by the decisions of the dean of a university, let alone an outside industry representative. “If you’re a faculty member in that unit you should know that the university is fine with auctioning off your academic freedom,” he said. “That’s what they’re doing.”

Ives of LSU said its Institute for Energy Innovation is no different to similar institutes across the US, including the Texas Bureau of Economic Geology, which performs research supported by corporate donors. “I think researchers saying that somehow having corporate funding for research damages the integrity of that research is a little far-fetched,” Ives said.

Research performed at the institute is subject to the faculty’s individual ethics training and subject to peer-review, he said. “A donor that provided money that goes to the institute isn’t going to be able to influence the outcome of that research in any way.”

Asked about the relationship with the institute and industry, Karsten Thompson, the interim dean of the College of Engineering at LSU said: “To me, it’s not a conflict at all. It’s a partnership because they’re the ones that are going to make the largest initial impacts on reducing CO 2 emissions.”

Some observers, noting that fossil fuel companies have previously shown a vested interest in obscuring scientific conclusions, question the reliability of academic studies sponsored by fossil fuel companies. Exxon, for example, denied the risk of human-caused climate change for decades , noted Jane Patton, an LSU alumna and the US fossil economy campaign manager for the Center for International Environmental Law.

After the Lens asked her to review LSU communication on the matter, Patton said she suspected that fossil fuel companies have had a say in what does and doesn’t get studied in relation to risky endeavors, such as carbon capture, which involves chemically stripping carbon dioxide from industrial emissions and piping it underground. For her, the LSU documents basically proved her fear. “This is the first time I’ve seen actual evidence of it,” Patton said. “This is a gross misuse of the public trust.”

To Patton, the perceived blurring of academic objectivity could not come at a worse time in Louisiana, as the climate crisis makes the state less habitable and housing more expensive . “It’s just disheartening,” she said, “to find that the state’s flagship institution is allowing industry to determine the research agenda. No wonder it’s so hard to find peer-reviewed research about how bad this is.”

Records show that Shell helped to tailor what LSU students would learn in the six courses offered under the institute’s carbon capture, use and storage (CCUS) concentration that debuted a couple years ago. The LSU alumnus Lee Stockwell, Shell’s general manager of CCUS, sat on the search committee for the Energy Institute executive director, served on the petroleum engineering advisory board, and was very involved in shaping the carbon capture curriculum.

Stockwell directed questions about Shell’s partnership with the university to LSU.

Stockwell was not the only oil representative to help design the curriculum. BP, Chevron, ConocoPhillips and ExxonMobil also had representatives on the ad hoc advisory committee that designed carbon capture coursework within the petroleum engineering department, according to a July 2022 email from Thompson . At least one cohort of students took two elective courses at LSU designed by the oil majors and another 10 students were expected to take the full concentration beginning in 2022.

LSU is not alone in this practice, Thompson said. At most engineering departments in the country, an active Industrial Advisory Committee (IAC) weighs in on curricula, so that degrees evolve as technology changes, helping students land internships and jobs.

LSU faculty has not been similarly engaged with renewable energy companies, because oil and gas companies have the resources to tackle the climate crisis now – and are not reliant on future technology, Thompson said. “Renewable energy is much more abstract,” he said. “So, I think that’s the difference. It’s not that we don’t care as much.”

Fossil fuel companies have been finding their way into classrooms for decades, in part to help the industry retain a positive public image in the face of a heating planet.

Some students do not approve of the university’s partnerships with fossil fuel companies, or any financial ties with them.

For a decade now, students across the nation have filed complaints and demanded divestment from fossil fuels and hundreds of institutions have agreed. Locally, the LSU Climate Pelicans, an interdisciplinary group of students, have called for the university to divest endowment funds from the fossil fuel industry.

Inspired by the Climate Pelicans’ work toward divestment, the LSU graduate student Alicia Cerquone, who sits on the LSU’s student senate, sponsored a divestment resolution. The measure passed in a 37-2 vote last year, according to LSU’s student newspaper . Though investment in fossil fuels amounts to only 2 to 3% of the endowment, it’s an important philosophical step, Cerquone said.

Cerquone is also troubled by the influence that industry has on the Institute for Energy Innovation and fears other corporations could control other departments’ curriculums. “These entities are going to have a say in what we pay to learn here,” she said.

The fossil fuel industry has made forays into academia beyond Louisiana. ExxonMobil and Shell have both helped fund a similar Energy Initiative at Massachusetts Institute of Technology (MIT), where the highest-level donors can have an office on MIT’s campus, according to Inside Climate News . In 2021, Exxon funded and co-wrote a research paper with MIT researchers with conclusions that supported the argument for federal subsidies for carbon capture and use.

This story is co published with the Lens , a non-profit newsroom in New Orleans and part of its Captured Audience series, which is supported by a grant from the Fund for Investigative Journalism

• Oil and gas companies
• Big oil uncovered
• US universities
• US education
• Climate crisis
• Energy industry

Most viewed