limitations of statistics in social work research

Social Work Research Methods That Drive the Practice

Social workers advocate for the well-being of individuals, families and communities. But how do social workers know what interventions are needed to help an individual? How do they assess whether a treatment plan is working? What do social workers use to write evidence-based policy?

Social work involves research-informed practice and practice-informed research. At every level, social workers need to know objective facts about the populations they serve, the efficacy of their interventions and the likelihood that their policies will improve lives. A variety of social work research methods make that possible.

Data-Driven Work

Data is a collection of facts used for reference and analysis. In a field as broad as social work, data comes in many forms.

Quantitative vs. Qualitative

As with any research, social work research involves both quantitative and qualitative studies.

Quantitative Research

Answers to questions like these can help social workers know about the populations they serve — or hope to serve in the future.

• How many students currently receive reduced-price school lunches in the local school district?
• How many hours per week does a specific individual consume digital media?
• How frequently did community members access a specific medical service last year?

Quantitative data — facts that can be measured and expressed numerically — are crucial for social work.

Quantitative research has advantages for social scientists. Such research can be more generalizable to large populations, as it uses specific sampling methods and lends itself to large datasets. It can provide important descriptive statistics about a specific population. Furthermore, by operationalizing variables, it can help social workers easily compare similar datasets with one another.

Qualitative Research

Qualitative data — facts that cannot be measured or expressed in terms of mere numbers or counts — offer rich insights into individuals, groups and societies. It can be collected via interviews and observations.

• What attitudes do students have toward the reduced-price school lunch program?
• What strategies do individuals use to moderate their weekly digital media consumption?
• What factors made community members more or less likely to access a specific medical service last year?

Qualitative research can thereby provide a textured view of social contexts and systems that may not have been possible with quantitative methods. Plus, it may even suggest new lines of inquiry for social work research.

Mixed Methods Research

Combining quantitative and qualitative methods into a single study is known as mixed methods research. This form of research has gained popularity in the study of social sciences, according to a 2019 report in the academic journal Theory and Society. Since quantitative and qualitative methods answer different questions, merging them into a single study can balance the limitations of each and potentially produce more in-depth findings.

However, mixed methods research is not without its drawbacks. Combining research methods increases the complexity of a study and generally requires a higher level of expertise to collect, analyze and interpret the data. It also requires a greater level of effort, time and often money.

The Importance of Research Design

Data-driven practice plays an essential role in social work. Unlike philanthropists and altruistic volunteers, social workers are obligated to operate from a scientific knowledge base.

To know whether their programs are effective, social workers must conduct research to determine results, aggregate those results into comprehensible data, analyze and interpret their findings, and use evidence to justify next steps.

Employing the proper design ensures that any evidence obtained during research enables social workers to reliably answer their research questions.

Research Methods in Social Work

The various social work research methods have specific benefits and limitations determined by context. Common research methods include surveys, program evaluations, needs assessments, randomized controlled trials, descriptive studies and single-system designs.

Surveys involve a hypothesis and a series of questions in order to test that hypothesis. Social work researchers will send out a survey, receive responses, aggregate the results, analyze the data, and form conclusions based on trends.

Surveys are one of the most common research methods social workers use — and for good reason. They tend to be relatively simple and are usually affordable. However, surveys generally require large participant groups, and self-reports from survey respondents are not always reliable.

Program Evaluations

Social workers ally with all sorts of programs: after-school programs, government initiatives, nonprofit projects and private programs, for example.

Crucially, social workers must evaluate a program’s effectiveness in order to determine whether the program is meeting its goals and what improvements can be made to better serve the program’s target population.

Evidence-based programming helps everyone save money and time, and comparing programs with one another can help social workers make decisions about how to structure new initiatives. Evaluating programs becomes complicated, however, when programs have multiple goal metrics, some of which may be vague or difficult to assess (e.g., “we aim to promote the well-being of our community”).

Needs Assessments

Social workers use needs assessments to identify services and necessities that a population lacks access to.

Common social work populations that researchers may perform needs assessments on include:

• People in a specific income group
• Everyone in a specific geographic region
• A specific ethnic group
• People in a specific age group

In the field, a social worker may use a combination of methods (e.g., surveys and descriptive studies) to learn more about a specific population or program. Social workers look for gaps between the actual context and a population’s or individual’s “wants” or desires.

For example, a social worker could conduct a needs assessment with an individual with cancer trying to navigate the complex medical-industrial system. The social worker may ask the client questions about the number of hours they spend scheduling doctor’s appointments, commuting and managing their many medications. After learning more about the specific client needs, the social worker can identify opportunities for improvements in an updated care plan.

In policy and program development, social workers conduct needs assessments to determine where and how to effect change on a much larger scale. Integral to social work at all levels, needs assessments reveal crucial information about a population’s needs to researchers, policymakers and other stakeholders. Needs assessments may fall short, however, in revealing the root causes of those needs (e.g., structural racism).

Randomized Controlled Trials

Randomized controlled trials are studies in which a randomly selected group is subjected to a variable (e.g., a specific stimulus or treatment) and a control group is not. Social workers then measure and compare the results of the randomized group with the control group in order to glean insights about the effectiveness of a particular intervention or treatment.

Randomized controlled trials are easily reproducible and highly measurable. They’re useful when results are easily quantifiable. However, this method is less helpful when results are not easily quantifiable (i.e., when rich data such as narratives and on-the-ground observations are needed).

Descriptive Studies

Descriptive studies immerse the researcher in another context or culture to study specific participant practices or ways of living. Descriptive studies, including descriptive ethnographic studies, may overlap with and include other research methods:

• Informant interviews
• Census data
• Observation

By using descriptive studies, researchers may glean a richer, deeper understanding of a nuanced culture or group on-site. The main limitations of this research method are that it tends to be time-consuming and expensive.

Single-System Designs

Unlike most medical studies, which involve testing a drug or treatment on two groups — an experimental group that receives the drug/treatment and a control group that does not — single-system designs allow researchers to study just one group (e.g., an individual or family).

Single-system designs typically entail studying a single group over a long period of time and may involve assessing the group’s response to multiple variables.

For example, consider a study on how media consumption affects a person’s mood. One way to test a hypothesis that consuming media correlates with low mood would be to observe two groups: a control group (no media) and an experimental group (two hours of media per day). When employing a single-system design, however, researchers would observe a single participant as they watch two hours of media per day for one week and then four hours per day of media the next week.

These designs allow researchers to test multiple variables over a longer period of time. However, similar to descriptive studies, single-system designs can be fairly time-consuming and costly.

Learn More About Social Work Research Methods

Social workers have the opportunity to improve the social environment by advocating for the vulnerable — including children, older adults and people with disabilities — and facilitating and developing resources and programs.

Learn more about how you can earn your  Master of Social Work online at Virginia Commonwealth University . The highest-ranking school of social work in Virginia, VCU has a wide range of courses online. That means students can earn their degrees with the flexibility of learning at home. Learn more about how you can take your career in social work further with VCU.

From M.S.W. to LCSW: Understanding Your Career Path as a Social Worker

How Palliative Care Social Workers Support Patients With Terminal Illnesses

How to Become a Social Worker in Health Care

Gov.uk, Mixed Methods Study

MVS Open Press, Foundations of Social Work Research

Open Social Work Education, Scientific Inquiry in Social Work

Open Social Work, Graduate Research Methods in Social Work: A Project-Based Approach

Routledge, Research for Social Workers: An Introduction to Methods

SAGE Publications, Research Methods for Social Work: A Problem-Based Approach

Theory and Society, Mixed Methods Research: What It Is and What It Could Be

READY TO GET STARTED WITH OUR ONLINE M.S.W. PROGRAM FORMAT?

Bachelor’s degree is required.

VCU Program Helper

This AI chatbot provides automated responses, which may not always be accurate. By continuing with this conversation, you agree that the contents of this chat session may be transcribed and retained. You also consent that this chat session and your interactions, including cookie usage, are subject to our privacy policy .

• Architecture and Design
• Asian and Pacific Studies
• Business and Economics
• Classical and Ancient Near Eastern Studies
• Computer Sciences
• Cultural Studies
• Engineering
• General Interest
• Geosciences
• Industrial Chemistry
• Islamic and Middle Eastern Studies
• Jewish Studies
• Library and Information Science, Book Studies
• Life Sciences
• Linguistics and Semiotics
• Literary Studies
• Materials Sciences
• Mathematics
• Social Sciences
• Sports and Recreation
• Theology and Religion
• Publish your article
• The role of authors
• Promoting your article
• Abstracting & indexing
• Publishing Ethics
• Why publish with De Gruyter
• How to publish with De Gruyter
• Our book series
• Our subject areas
• Your digital product at De Gruyter
• Contribute to our reference works
• Product information
• Tools & resources
• Product Information
• Promotional Materials
• Orders and Inquiries
• FAQ for Library Suppliers and Book Sellers
• Repository Policy
• Free access policy
• Open Access agreements
• Database portals
• For Authors
• Customer service
• People + Culture
• Journal Management
• How to join us
• Working at De Gruyter
• Mission & Vision
• De Gruyter Foundation
• De Gruyter Ebound
• Our Responsibility
• Partner publishers

Your purchase has been completed. Your documents are now available to view.

Statistics in Social Work

An introduction to practical applications.

• Amy Batchelor
• X / Twitter

Please login or register with De Gruyter to order this product.

• Language: English
• Publisher: Columbia University Press
• Copyright year: 2019
• Audience: Professional and scholarly;
• Other: 50 figures
• Published: November 19, 2019
• ISBN: 9780231550222

• Search Menu
• Advance articles
• Editor's Choice
• Author Guidelines
• Submission Site
• Open Access
• About The British Journal of Social Work
• About the British Association of Social Workers
• Editorial Board
• Advertising and Corporate Services
• Journals Career Network
• Self-Archiving Policy
• Dispatch Dates
• Journals on Oxford Academic
• Books on Oxford Academic

• < Previous

Quantitative Research Methods for Social Work: Making Social Work Count, Barbra Teater, John Devaney, Donald Forester, Jonathan Scourfield and John Carpenter

• Article contents
• Figures & tables
• Supplementary Data

Hugh McLaughlin, Quantitative Research Methods for Social Work: Making Social Work Count, Barbra Teater, John Devaney, Donald Forester, Jonathan Scourfield and John Carpenter, The British Journal of Social Work , Volume 52, Issue 3, April 2022, Pages 1793–1795, https://doi.org/10.1093/bjsw/bcaa116

• Permissions Icon Permissions

I remember sharing a lift with a Professor from America at a joint IFSW/IASSW World Social Work Conference who taught research methods. After chatting about teaching research methods, he informed me gleefully that his students are taught qualitative methods first. However, after they get to him, none of them leave his classroom without being ‘converted to quantitative methods’! At this point, he got off the lift leaving our discussion in the air.

This book arose from funding from the Economic and Social Research Council to address the quantitative skills gap in the social sciences. The grants were applied for under the auspices of the Joint University Council Social Work Education Committee to upskill social work academics and develop a curriculum resource with teaching aids. I was saddened to discover that many of the free resources are no longer available and wondered if anything could be done to remedy this.

The book is unusual for the UK in that its major focus is on quantitative methods unlike other social work research methods books which tend to cover both qualitative and quantitative methods ( Campbell et al. , 2017; Smith, 2009). Until this book came along many of us will have been happy using non-social work research methods texts to learn about quantitative methods ( Bryman, 2015). This authoritative text offers a fresh and imaginative approach to teaching quantitative methods. It is set out in an incremental and easily accessible format with a series of exercises and critical thinking boxes with suggested readings at the end of each chapter. The exercises and critical thinking questions are well thought out with a full answer to each at the back of the book—thus making it really useful for those of us who teach research methods. It is also aimed at social work academics, social work students and practitioners who want to learn more about quantitative approaches, where they are useful, how they can be read and understood, and how they can be applied to their setting.

Email alerts

Citing articles via.

• Recommend to your Library

Affiliations

• Online ISSN 1468-263X
• Print ISSN 0045-3102
• Copyright © 2024 British Association of Social Workers
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

(Stanford users can avoid this Captcha by logging in.)

• Send to text email RefWorks EndNote printer

Statistics in social work : an introduction to practical applications

Available online.

• EBSCO University Press

More options

• Find it at other libraries via WorldCat
• Contributors

Description

Creators/contributors, contents/summary.

• Acknowledgments 1. Introduction 2. Creating Useful Data 3. Understanding People and Populations 4. Variance: The Distance Between Us 5. The Statistics of Relationships 6. Sampling: The Who and the How 7. What Works? Hypothesis Testing and Inferential Statistics 8. When Two Is Not Enough: Testing with Multiple Groups 9. An Introduction to Advanced Concepts
• Appendix I: Glossary
• Appendix II: Answer Key for Review Questions
• Appendix III: Equations Cheat Sheet References Index.
• (source: Nielsen Book Data)

Bibliographic information

Browse related items.

• Stanford Home
• Maps & Directions
• Search Stanford
• Emergency Info
• Terms of Use
• Non-Discrimination
• Accessibility

© Stanford University , Stanford , California 94305 .

The Strengths and Limitations of Social Work

• First Online: 15 November 2023

Cite this chapter

• Fiona McDermott   ORCID: orcid.org/0000-0002-9362-1441 6 ,
• Kerry Brydon   ORCID: orcid.org/0000-0003-4373-8112 7 ,
• Alex Haynes 8 &
• Felicity Moon   ORCID: orcid.org/0000-0002-0317-8598 9  

161 Accesses

The focus of this chapter is on describing the strengths of social work in order to present the case for building upon these strengths as social work continues to evolve and adapt in a world at the beginning of the twenty-first century, which has altered in so many ways due to such profound influences as advances in telecommunications and social media, climate change, the COVID-19 pandemic, mass migrations, the Russian invasion of Ukraine, to name just a few. These strengths are explored. This is followed by discussion of some limitations of current social work theorising particularly in relation to ‘wicked problems’; social workers’ responses to risk and uncertainty; the weaknesses of current static or fixed depictions of social reality at micro, meso and macro ‘levels’; and the inability of person-in-environment formulations to explain how exactly person and environment intersect and mutually influence each other. This critique provides the springboard for a discussion of complexity theory, which will be more fully explored in subsequent chapters.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

• Available as EPUB and PDF
• Read on any device
• Instant download
• Own it forever
• Durable hardcover edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

AASW. (2020). Code of ethics https://www.aasw.asn.au/document/item/92

Asquith, S., Clark, C., & Waterhouse, L. (2005). The role of the social worker in the 21st century – A literature review, Edinburgh: Scottish Executive Education Department. Retrieved from: https://www.webarchive.org.uk/wayback/archive/3000/https://www.gov.scot/Resource/Doc/47121/0020821.pdf

Australian Institute of Health and Welfare. (2022). Social determinants of health . https://www.aihw.gov.au/reports/australias-health/social-determinants-of-health

Cameron, N., & McDermott, F. (2007). Social work and the body . Palgrave Macmillan.

Book   Google Scholar  

Cree, V., & Wallace, S. (2009). Risk and protection. In R. Adams, L. Dominelli, & M. Payne (Eds.), Practising social work in a complex world (2nd ed., pp. 42–56). Palgrave Macmillan.

Chapter   Google Scholar  

Davies, M. (1994). The essential social worker: An introduction to professional practice in the 1990s (3rd ed.). Ashgate.

Google Scholar  

Dominelli, L. (2002). Anti-oppressive social work theory and practice . Palgrave.

Evans, T., & Hardy, M. (2010). Skills in contemporary social work . Polity Press.

Flack, J., & Mitchell, M. (2020), Uncertain times: The pandemic is an unprecedented opportunity – Seeing human society as a complex systems opens a better future for us all, Aeon Essay . Retrieved from: https://aeon.co/essays/complex-systems-science-allows-us-to-see-new-paths-forward .

Furlong, M. (2013). Building the client’s relational base . The Policy Press.

Gitterman, A., & Germain, C. B. (2008). The life model of social work practice: Advances in knowledge and practice (3rd ed.). Columbia University Press.

Glass, D. (2021). The Ombudsman for human rights: A case book , Melbourne: Victorian Government Printer. Retrieved from: https://assets.ombudsman.vic.gov.au/assets/The-Ombudsman-for-Human-Rights-A-Casebook-Aug-2021.pdf .

Grant, S. (2021). With the falling of dusk . Harper Collins.

Hare, I. (2004). Defining social work for the 21st century: The International Federation of Social Workers revised definition of social work. International Social Work, 47 (3), 407–424. https://doi.org/10.1177/0020872804043973

Article   Google Scholar  

Howe, D. (2014). The compleat social worker . Palgrave.

International Federation of Social Workers. (2014). Global definition of social work. International Federation of Social Workers. Retrieved from: https://www.ifsw.org/what-is-social-work/global-definition-of-social-work/ .

International Federation of Social Workers. (2018). Global social work statement of ethical principles . https://www.ifsw.org/global-social-work-statement-of-ethical-principles/

Jackson, M. (2019). Critical systems thinking and the management of complexity . Wiley.

McDermott, F. (2014). Complexity theory, transdisciplinary working and reflective practice. In A. Pycroft & C. Bartollas (Eds.), Applying complexity theory . Policy Press.

McDermott, F., Henderson, A., & Quayle, C. (2017). Health social workers’ sources of knowledge for decision making in practice. Social Work in Health Care, 56 , 794. https://doi.org/10.1080/00981389.2017.1340391

Milner, J., Myers, S., & O’Byrne, P. (2020). Assessment in social work (5th ed.). Macmillan Education Ltd.

Payne, M. (2014). Modern social work theory (4th ed.). Lyceum Books.

Peters, B. (2017). What is so wicked about wicked problems? A conceptual analysis and a research program. Policy and Society, 36 (3), 385–396. https://doi.org/10.1080/14494035.2017.1361633

Rittell, H. W. J. & Webber, M. M. (1973). Dilemmas in a general theory of planning. Policy Sciences, 4 , 155–169.

Rollins, W. (2019). Social worker-client relationships: Social worker perspectives. Australian Social Work, 73 (4), 395–407. https://doi.org/10.1080/0312407X.2019.1669687

Schon, D. (1992). The reflective practitioner: How professionals think in action . Routledge.

Scoones, I. (2019). What is uncertainty and why does it matter . STEPS Centre Working Paper 105. Retrieved from: https://opendocs.ids.ac.uk/opendocs/bitstream/handle/20.500.12413/14470/STEPSWP_105_Scoones_final.pdf?sequence=1&isAllowed=y

Smith, J., Puckett, C., & Simon, W. (2016). Indigenous allyship: An overview . Wilfred Laurier University.

Social Determinants of Health. (n.d.). https://www.aihw.gov.au/reports/australias-health/social-determinants-of-health

Termeer, C., Dewulf, A., & Biesbroek, R. (2019). A critical assessment of the wicked problem concept and usefulness for policy science and practice. Policy and Society, 38 (2), 167–179. https://doi.org/10.1080/14494035.2019.1617971

Warren, K., Franklin, C., & Streeter, C. (1998). New directions in systems theory: Chaos and complexity. Social Work, 43 (4), 357–372.

Watts, L. (2021). Values, beliefs, and attitudes about reflective practice in Australian social work education and practice. Australian Social Work , 1. https://doi.org/10.1080/0312407X.2021.1874031

Wolf-Branigin, M. (2009). Applying complexity and emergence in social work education. Social Work Education, 28 (2), 115–127. https://doi.org/10.1080/02615470802028090

Zajda, J., Majhanovich, S., & Rust, V. (2006). Introduction: Education and social justice. International Review of Education, 52 (1), 9–22. https://doi.org/10.1007/s11159-005-5614-2

Download references

Author information

Authors and affiliations.

Department of Social Work, Faculty of Medicine, Nursing & Health Sciences, School of Primary & Allied Health Care, Monash University, Caulfield East, VIC, Australia

Fiona McDermott

Social Work Practitioner, Melbourne, VIC, Australia

Kerry Brydon

Whittlesea Community Connections Inc., Epping, VIC, Australia

Alex Haynes

Emergency Department, Royal Melbourne Hospital, Parkville, VIC, Australia

Felicity Moon

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Fiona McDermott .

Rights and permissions

Reprints and permissions

Copyright information

© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this chapter

McDermott, F., Brydon, K., Haynes, A., Moon, F. (2024). The Strengths and Limitations of Social Work. In: Complexity Theory for Social Work Practice. Springer, Cham. https://doi.org/10.1007/978-3-031-38677-0_2

Download citation

DOI : https://doi.org/10.1007/978-3-031-38677-0_2

Published : 15 November 2023

Publisher Name : Springer, Cham

Print ISBN : 978-3-031-38676-3

Online ISBN : 978-3-031-38677-0

eBook Packages : Social Sciences Social Sciences (R0)

Share this chapter

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

• Publish with us

Policies and ethics

• Find a journal
• Track your research

• USC Libraries
• Research Guides

Organizing Your Social Sciences Research Paper

• Limitations of the Study
• Purpose of Guide
• Design Flaws to Avoid
• Independent and Dependent Variables
• Glossary of Research Terms
• Reading Research Effectively
• Narrowing a Topic Idea
• Broadening a Topic Idea
• Extending the Timeliness of a Topic Idea
• Academic Writing Style
• Applying Critical Thinking
• Choosing a Title
• Making an Outline
• Paragraph Development
• Research Process Video Series
• Executive Summary
• The C.A.R.S. Model
• Background Information
• The Research Problem/Question
• Theoretical Framework
• Citation Tracking
• Content Alert Services
• Evaluating Sources
• Primary Sources
• Secondary Sources
• Tiertiary Sources
• Scholarly vs. Popular Publications
• Qualitative Methods
• Quantitative Methods
• Insiderness
• Using Non-Textual Elements
• Common Grammar Mistakes
• Writing Concisely
• Avoiding Plagiarism
• Footnotes or Endnotes?
• Further Readings
• Generative AI and Writing
• USC Libraries Tutorials and Other Guides
• Bibliography

The limitations of the study are those characteristics of design or methodology that impacted or influenced the interpretation of the findings from your research. Study limitations are the constraints placed on the ability to generalize from the results, to further describe applications to practice, and/or related to the utility of findings that are the result of the ways in which you initially chose to design the study or the method used to establish internal and external validity or the result of unanticipated challenges that emerged during the study.

Price, James H. and Judy Murnan. “Research Limitations and the Necessity of Reporting Them.” American Journal of Health Education 35 (2004): 66-67; Theofanidis, Dimitrios and Antigoni Fountouki. "Limitations and Delimitations in the Research Process." Perioperative Nursing 7 (September-December 2018): 155-163. .

Importance of...

Always acknowledge a study's limitations. It is far better that you identify and acknowledge your study’s limitations than to have them pointed out by your professor and have your grade lowered because you appeared to have ignored them or didn't realize they existed.

Keep in mind that acknowledgment of a study's limitations is an opportunity to make suggestions for further research. If you do connect your study's limitations to suggestions for further research, be sure to explain the ways in which these unanswered questions may become more focused because of your study.

Acknowledgment of a study's limitations also provides you with opportunities to demonstrate that you have thought critically about the research problem, understood the relevant literature published about it, and correctly assessed the methods chosen for studying the problem. A key objective of the research process is not only discovering new knowledge but also to confront assumptions and explore what we don't know.

Claiming limitations is a subjective process because you must evaluate the impact of those limitations . Don't just list key weaknesses and the magnitude of a study's limitations. To do so diminishes the validity of your research because it leaves the reader wondering whether, or in what ways, limitation(s) in your study may have impacted the results and conclusions. Limitations require a critical, overall appraisal and interpretation of their impact. You should answer the question: do these problems with errors, methods, validity, etc. eventually matter and, if so, to what extent?

Price, James H. and Judy Murnan. “Research Limitations and the Necessity of Reporting Them.” American Journal of Health Education 35 (2004): 66-67; Structure: How to Structure the Research Limitations Section of Your Dissertation. Dissertations and Theses: An Online Textbook. Laerd.com.

Descriptions of Possible Limitations

All studies have limitations . However, it is important that you restrict your discussion to limitations related to the research problem under investigation. For example, if a meta-analysis of existing literature is not a stated purpose of your research, it should not be discussed as a limitation. Do not apologize for not addressing issues that you did not promise to investigate in the introduction of your paper.

Here are examples of limitations related to methodology and the research process you may need to describe and discuss how they possibly impacted your results. Note that descriptions of limitations should be stated in the past tense because they were discovered after you completed your research.

Possible Methodological Limitations

• Sample size -- the number of the units of analysis you use in your study is dictated by the type of research problem you are investigating. Note that, if your sample size is too small, it will be difficult to find significant relationships from the data, as statistical tests normally require a larger sample size to ensure a representative distribution of the population and to be considered representative of groups of people to whom results will be generalized or transferred. Note that sample size is generally less relevant in qualitative research if explained in the context of the research problem.
• Lack of available and/or reliable data -- a lack of data or of reliable data will likely require you to limit the scope of your analysis, the size of your sample, or it can be a significant obstacle in finding a trend and a meaningful relationship. You need to not only describe these limitations but provide cogent reasons why you believe data is missing or is unreliable. However, don’t just throw up your hands in frustration; use this as an opportunity to describe a need for future research based on designing a different method for gathering data.
• Lack of prior research studies on the topic -- citing prior research studies forms the basis of your literature review and helps lay a foundation for understanding the research problem you are investigating. Depending on the currency or scope of your research topic, there may be little, if any, prior research on your topic. Before assuming this to be true, though, consult with a librarian! In cases when a librarian has confirmed that there is little or no prior research, you may be required to develop an entirely new research typology [for example, using an exploratory rather than an explanatory research design ]. Note again that discovering a limitation can serve as an important opportunity to identify new gaps in the literature and to describe the need for further research.
• Measure used to collect the data -- sometimes it is the case that, after completing your interpretation of the findings, you discover that the way in which you gathered data inhibited your ability to conduct a thorough analysis of the results. For example, you regret not including a specific question in a survey that, in retrospect, could have helped address a particular issue that emerged later in the study. Acknowledge the deficiency by stating a need for future researchers to revise the specific method for gathering data.
• Self-reported data -- whether you are relying on pre-existing data or you are conducting a qualitative research study and gathering the data yourself, self-reported data is limited by the fact that it rarely can be independently verified. In other words, you have to the accuracy of what people say, whether in interviews, focus groups, or on questionnaires, at face value. However, self-reported data can contain several potential sources of bias that you should be alert to and note as limitations. These biases become apparent if they are incongruent with data from other sources. These are: (1) selective memory [remembering or not remembering experiences or events that occurred at some point in the past]; (2) telescoping [recalling events that occurred at one time as if they occurred at another time]; (3) attribution [the act of attributing positive events and outcomes to one's own agency, but attributing negative events and outcomes to external forces]; and, (4) exaggeration [the act of representing outcomes or embellishing events as more significant than is actually suggested from other data].

Possible Limitations of the Researcher

• Access -- if your study depends on having access to people, organizations, data, or documents and, for whatever reason, access is denied or limited in some way, the reasons for this needs to be described. Also, include an explanation why being denied or limited access did not prevent you from following through on your study.
• Longitudinal effects -- unlike your professor, who can literally devote years [even a lifetime] to studying a single topic, the time available to investigate a research problem and to measure change or stability over time is constrained by the due date of your assignment. Be sure to choose a research problem that does not require an excessive amount of time to complete the literature review, apply the methodology, and gather and interpret the results. If you're unsure whether you can complete your research within the confines of the assignment's due date, talk to your professor.
• Cultural and other type of bias -- we all have biases, whether we are conscience of them or not. Bias is when a person, place, event, or thing is viewed or shown in a consistently inaccurate way. Bias is usually negative, though one can have a positive bias as well, especially if that bias reflects your reliance on research that only support your hypothesis. When proof-reading your paper, be especially critical in reviewing how you have stated a problem, selected the data to be studied, what may have been omitted, the manner in which you have ordered events, people, or places, how you have chosen to represent a person, place, or thing, to name a phenomenon, or to use possible words with a positive or negative connotation. NOTE :   If you detect bias in prior research, it must be acknowledged and you should explain what measures were taken to avoid perpetuating that bias. For example, if a previous study only used boys to examine how music education supports effective math skills, describe how your research expands the study to include girls.
• Fluency in a language -- if your research focuses , for example, on measuring the perceived value of after-school tutoring among Mexican-American ESL [English as a Second Language] students and you are not fluent in Spanish, you are limited in being able to read and interpret Spanish language research studies on the topic or to speak with these students in their primary language. This deficiency should be acknowledged.

Aguinis, Hermam and Jeffrey R. Edwards. “Methodological Wishes for the Next Decade and How to Make Wishes Come True.” Journal of Management Studies 51 (January 2014): 143-174; Brutus, Stéphane et al. "Self-Reported Limitations and Future Directions in Scholarly Reports: Analysis and Recommendations." Journal of Management 39 (January 2013): 48-75; Senunyeme, Emmanuel K. Business Research Methods. Powerpoint Presentation. Regent University of Science and Technology; ter Riet, Gerben et al. “All That Glitters Isn't Gold: A Survey on Acknowledgment of Limitations in Biomedical Studies.” PLOS One 8 (November 2013): 1-6.

Structure and Writing Style

Information about the limitations of your study are generally placed either at the beginning of the discussion section of your paper so the reader knows and understands the limitations before reading the rest of your analysis of the findings, or, the limitations are outlined at the conclusion of the discussion section as an acknowledgement of the need for further study. Statements about a study's limitations should not be buried in the body [middle] of the discussion section unless a limitation is specific to something covered in that part of the paper. If this is the case, though, the limitation should be reiterated at the conclusion of the section.

If you determine that your study is seriously flawed due to important limitations , such as, an inability to acquire critical data, consider reframing it as an exploratory study intended to lay the groundwork for a more complete research study in the future. Be sure, though, to specifically explain the ways that these flaws can be successfully overcome in a new study.

But, do not use this as an excuse for not developing a thorough research paper! Review the tab in this guide for developing a research topic . If serious limitations exist, it generally indicates a likelihood that your research problem is too narrowly defined or that the issue or event under study is too recent and, thus, very little research has been written about it. If serious limitations do emerge, consult with your professor about possible ways to overcome them or how to revise your study.

When discussing the limitations of your research, be sure to:

• Describe each limitation in detailed but concise terms;
• Explain why each limitation exists;
• Provide the reasons why each limitation could not be overcome using the method(s) chosen to acquire or gather the data [cite to other studies that had similar problems when possible];
• Assess the impact of each limitation in relation to the overall findings and conclusions of your study; and,
• If appropriate, describe how these limitations could point to the need for further research.

Remember that the method you chose may be the source of a significant limitation that has emerged during your interpretation of the results [for example, you didn't interview a group of people that you later wish you had]. If this is the case, don't panic. Acknowledge it, and explain how applying a different or more robust methodology might address the research problem more effectively in a future study. A underlying goal of scholarly research is not only to show what works, but to demonstrate what doesn't work or what needs further clarification.

Aguinis, Hermam and Jeffrey R. Edwards. “Methodological Wishes for the Next Decade and How to Make Wishes Come True.” Journal of Management Studies 51 (January 2014): 143-174; Brutus, Stéphane et al. "Self-Reported Limitations and Future Directions in Scholarly Reports: Analysis and Recommendations." Journal of Management 39 (January 2013): 48-75; Ioannidis, John P.A. "Limitations are not Properly Acknowledged in the Scientific Literature." Journal of Clinical Epidemiology 60 (2007): 324-329; Pasek, Josh. Writing the Empirical Social Science Research Paper: A Guide for the Perplexed. January 24, 2012. Academia.edu; Structure: How to Structure the Research Limitations Section of Your Dissertation. Dissertations and Theses: An Online Textbook. Laerd.com; What Is an Academic Paper? Institute for Writing Rhetoric. Dartmouth College; Writing the Experimental Report: Methods, Results, and Discussion. The Writing Lab and The OWL. Purdue University.

Writing Tip

Don't Inflate the Importance of Your Findings!

After all the hard work and long hours devoted to writing your research paper, it is easy to get carried away with attributing unwarranted importance to what you’ve done. We all want our academic work to be viewed as excellent and worthy of a good grade, but it is important that you understand and openly acknowledge the limitations of your study. Inflating the importance of your study's findings could be perceived by your readers as an attempt hide its flaws or encourage a biased interpretation of the results. A small measure of humility goes a long way!

Another Writing Tip

Negative Results are Not a Limitation!

Negative evidence refers to findings that unexpectedly challenge rather than support your hypothesis. If you didn't get the results you anticipated, it may mean your hypothesis was incorrect and needs to be reformulated. Or, perhaps you have stumbled onto something unexpected that warrants further study. Moreover, the absence of an effect may be very telling in many situations, particularly in experimental research designs. In any case, your results may very well be of importance to others even though they did not support your hypothesis. Do not fall into the trap of thinking that results contrary to what you expected is a limitation to your study. If you carried out the research well, they are simply your results and only require additional interpretation.

Lewis, George H. and Jonathan F. Lewis. “The Dog in the Night-Time: Negative Evidence in Social Research.” The British Journal of Sociology 31 (December 1980): 544-558.

Yet Another Writing Tip

Sample Size Limitations in Qualitative Research

Sample sizes are typically smaller in qualitative research because, as the study goes on, acquiring more data does not necessarily lead to more information. This is because one occurrence of a piece of data, or a code, is all that is necessary to ensure that it becomes part of the analysis framework. However, it remains true that sample sizes that are too small cannot adequately support claims of having achieved valid conclusions and sample sizes that are too large do not permit the deep, naturalistic, and inductive analysis that defines qualitative inquiry. Determining adequate sample size in qualitative research is ultimately a matter of judgment and experience in evaluating the quality of the information collected against the uses to which it will be applied and the particular research method and purposeful sampling strategy employed. If the sample size is found to be a limitation, it may reflect your judgment about the methodological technique chosen [e.g., single life history study versus focus group interviews] rather than the number of respondents used.

Boddy, Clive Roland. "Sample Size for Qualitative Research." Qualitative Market Research: An International Journal 19 (2016): 426-432; Huberman, A. Michael and Matthew B. Miles. "Data Management and Analysis Methods." In Handbook of Qualitative Research . Norman K. Denzin and Yvonna S. Lincoln, eds. (Thousand Oaks, CA: Sage, 1994), pp. 428-444; Blaikie, Norman. "Confounding Issues Related to Determining Sample Size in Qualitative Research." International Journal of Social Research Methodology 21 (2018): 635-641; Oppong, Steward Harrison. "The Problem of Sampling in qualitative Research." Asian Journal of Management Sciences and Education 2 (2013): 202-210.

• << Previous: 8. The Discussion
• Next: 9. The Conclusion >>
• Last Updated: May 9, 2024 11:05 AM
• URL: https://libguides.usc.edu/writingguide

The use of statistics in social sciences

Journal of Humanities and Applied Social Sciences

ISSN : 2632-279X

Article publication date: 17 October 2019

Issue publication date: 30 October 2019

The purpose this paper is to review some of the statistical methods used in the field of social sciences.

Design/methodology/approach

A review of some of the statistical methodologies used in areas like survey methodology, official statistics, sociology, psychology, political science, criminology, public policy, marketing research, demography, education and economics.

Several areas are presented such as parametric modeling, nonparametric modeling and multivariate methods. Focus is also given to time series modeling, analysis of categorical data and sampling issues and other useful techniques for the analysis of data in the social sciences. Indicative references are given for all the above methods along with some insights for the application of these techniques.

Originality/value

This paper reviews some statistical methods that are used in social sciences and the authors draw the attention of researchers on less popular methods. The purpose is not to give technical details and also not to refer to all the existing techniques or to all the possible areas of statistics. The focus is mainly on the applied aspect of the techniques and the authors give insights about techniques that can be used to answer problems in the abovementioned areas of research.

• Time series
• Social sciences
• Non-parametric
• Multivariate
• Statistical modeling

Maravelakis, P. (2019), "The use of statistics in social sciences", Journal of Humanities and Applied Social Sciences , Vol. 1 No. 2, pp. 87-97. https://doi.org/10.1108/JHASS-08-2019-0038

Emerald Publishing Limited

Copyright © 2019, Petros Maravelakis.

Published in Journal of Humanities and Applied Social Sciences . Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode .

1. Introduction

According to Dodge (2008) , Statistics:

[…] is made up of a set of techniques for obtaining knowledge from incomplete data, from a rigorous scientific system for managing data collection, their organization, analysis, and interpretation, when it is possible to present them in numeric form.

The purpose of this paper is to try to review the statistical techniques in the field of social sciences in other words social statistics.

Social statistics is the field of statistical science that deals with the study of social phenomena and in particular human behavior in a social environment. Such phenomena are any kind of human activities, including activities of groups of people like households, societies and nations and their impacts on culture, education and other areas. Generally, we can say that social statistics deal with the application of statistical methodology in areas like survey methodology, official statistics, sociology, psychology, political science, criminology, public policy, marketing research, demography, education, economics and others.

Due to the nature of social sciences it is common to study indicators that cannot be measured directly. Moreover, data that is unobservable, informal, illegal or “too personal” are often studied in this area ( Lovric (2011) ). For example, a social researcher may be interested on the data (answers) of the question “Do you participate in illegal gambling?”. Other similar questions may be asking, for example, about the sexual behavior of a respondent, possible addictions etc.

In this paper, we review some statistical methods that are used in social sciences and we draw the attention of researchers on less popular methods. Our purpose is not to give technical details and also not to refer to all the existing techniques or to all the possible areas of statistics. We focus mainly on the applied aspect of the techniques and we give insights about techniques that can be used to answer problems in the abovementioned areas of research.

The remaining of the paper is organized as follows. In Section 2, we refer to the sampling issues which are of fundamental importance to gather data that are representative of the population under study. Section 3 gives an overview of descriptive statistics, regression analysis and analysis of variance. Since regression is a method heavily used to model quantitative data it is of great importance in social sciences. Analysis of variance can be used to identify if one or more categorical variables have a statistically significant effect on a continuous (dependent) variable. Parametric models are presented in Section 4. It is practically impossible to cover all the available models but a number of important models are outlined. Non-parametric methods are given in Section 5. These methods are particularly important because assumptions in parametric models are frequently questionable in practice. Some multivariate techniques are described in Section 6. Since most of the data contain several different variables that are related multivariate techniques are a necessary tool. Usual practice in questionnaires is to collect categorical data with inherent order and without order. In Section 7, we describe some of the popular techniques that analyze such data. Some of the time series methods to analyze data that are dependent over time are given in Section 8. In Section 9, we present the data mining methods that can be used to identify patterns in large data sets. Finally, we give some conclusions and recommendations.

2. Sampling

The first issue before the use of any statistical method is the collection of the sample. We may say that sampling is a collection of techniques on how to select a number of individuals from the target population to estimate certain characteristics of the population that we want to study. There are two ways to select a sample, either using a probability or a nonprobability sample. In probability sampling every unit of the population has a chance of being selected in the sample. Moreover, this probability is greater than zero and it can be computed. The methods that are used are simple random sampling, systematic sampling, stratified sampling, cluster sampling and multistage sampling (or combinations of these methods). On the other hand, in nonprobability sampling some of the units of the population have zero probability of being selected or the probability of being selected cannot be computed. The most known non-probability sampling methods are intentional sampling, convenience sampling, quota sampling and snowball sampling.

The main difference between probability and non-probability sampling is the fact that with probability sampling we draw a random sample. This assertion is extremely important due to the fact that using statistical theory we can extend the results from a random sample to draw conclusions about the population. This is not allowed in non-probability sampling.

The way to select a random sample is apparently very important. However one may observe that in papers published in social sciences this fact is overlooked. Specifically, emphasis is given to the conclusions of the research effort and not on the way data was gathered. This fact is surely a very important factor on the reliability of the results. In the same issue, another statistical question that needs special attention is the determination of the sample size. In many published papers the authors just refer to the size of the sample without referring to the technical details of its computation. This is another serious problem that affects the credibility of the results in a survey.

These two problems are not the only that appear in surveys in social sciences but surely they are the most usual. Other issues that occur are blurred definition of the population, problems during the collection of the data that are rarely mentioned (for example replacement of selected units), non-sampling errors (for example non-response, over or under coverage) and others.

Generally, we can say that in many cases sampling is treated with less attention than what is needed. Researchers in the field and practitioners may refer to the classical book by Kish (1995) . Other useful references are Cohen (1988) , Joseph et al. (1997) , Lenth (2001) and Shuster (1990) .

3. Descriptive statistics, regression and analysis of variance

The first step in any statistical analysis is the use of descriptive statistics to present the data and try to identify any kind of trends, relationships or abnormal behavior. Analysis based on descriptive statistics or exploratory data analysis usually makes no stochastic assumptions. A first approach in parametric tests is to use the classic hypothesis tests and confidence intervals. Apart from that there are also other statistical methods that can be employed in social sciences. Regression analysis and analysis of variance (ANOVA) are some of the classical methods.

3.1 Regression

Regression is one of the most known methods used for analyzing relationships between variables. The main objectives of a regression analysis is to check if there is an association between variables, to identify the strength of this relationship and to conclude to a regression equation that is used to describe this relationship.

There are several forms of regression modeling, for example, linear regression, logistic regression and regression discontinuity. There are also other aspects of the regression methodology but we confine ourselves to these cases. All these methodologies have been extensively used in real cases of social sciences.

Linear regression is the simplest of these methods since it is used to model the relationship between one dependent variable and one or more explanatory variables. In this methodology, we try to find a function to fit the values of the explanatory variable that vary linearly with the target variable. Linear regression is particularly useful since it is able to predict the value of the dependent variable given the value of the explanatory variable or variables. We have to stress that in this method the target variable (dependent variable) is continuous.

In logistic regression we want to obtain a nonlinear curve to fit the data when the target variable is discrete. This methodology is particularly useful in modeling a target variable having value for example Yes (0) or No (1). More formally we can say that the target variable is binomial. Our aim is to find an equation that functionally connects the values of the explanatory variables to the values of the target variable. The explanatory variables can be either continuous or categorical. Since the range of the explanatory variables can be between -∞ and ∞ a proper transformation is applied to the target variable. If we transform the target variable to the logarithm of the odds of its values then the transformed target variable is linearly related to the explanatory variables. For more details about the linear and the logistic regression the interested reader can refer to Kutner et al. (2005) .

Regression discontinuity is used to compute the effect of an intervention. This methodology is able to give unbiased estimates of this intervention. In a regression discontinuity design we use a rule to assign the intervention to a unit. This methodology is extensively used in education. Specifically, a scoring rule is used after a test is given in a class to select the students that need more effort on the specific course. Students with scores below a cutoff value are assigned to the group that will spend more time studying and students with scores above the cutoff value are assigned to the comparison group, or vice versa.

The effect of the intervention is estimated as the difference in the mean outcome of the treatment group and the comparison group. A regression line or curve is estimated for the two groups (treatment and comparison groups), and the difference in the mean of these regression lines at the cutoff value of the measured variable is the estimate of the effect of the intervention. We conclude that there was an effect of the intervention if a “discontinuity” appears between the two regression lines at the cutoff value. A detailed description of regression discontinuity is given in Riley-Tilman and Burns (2009) and Jacob and Zhu (2012) .

3.2 Analysis of variance

Analysis of variance (ANOVA) is a well-known method used to compare several means at the same time using a fixed confidence level. The data used are the results of an experiment. There is a continuous dependent variable, and one or more qualitative independent variables (categorical or nominal variables). The design of the experiment must be done in such a way that it will not affect its results. For example, a completely randomized experiment does not affect the output of the experiment. However, the choice of the design of the experiment affects which analysis of variance method will be used. There are a lot of different designs of experiments and analysis of variance methods for several different cases.

Regression analysis and analysis of variance are closely related. If we use dummy variables as independent variables in analysis of variance then the analysis becomes regression analysis. However, there is a serious difference between the two methods. In the analysis of variance if the design of the experiment is properly done, we may conclude that there is causality (the independent variable has a causal effect on the dependent variable). On the other hand, in regression analysis a statistically significant effect may mean causality or not (a statistically significant result does not necessarily mean causal effect).

The analysis of variance tests the independence of the response and explanatory variables. If we decide that there is this type of dependence then we have to do extra analysis to identify which means are different and to what extent.

The analysis of variance assumes that the samples in the groups (categories of the independent variable) are independent. This means that each group has a different sample of subjects. However, there are cases where each group has the same sample of subjects. Apparently, the samples are then dependent and of course we have to take this fact into consideration to reach credible results. This case is called repeated measures analysis of variance.

For more information on this topic, see Agresti and Finlay (2009) and Cohen and Lea (2004) .

4. Parametric methods

Assume that a researcher wants to use the ANOVA and apart from the dependent variable and a categorical variable (factor), data for one or more quantitative variables measured on each experimental unit are available. Then, if these variables have an effect on the outcome of the experiment, they can be used in the model as independent variables. Such variables are called covariates or concomitant variables. The analysis involving all these variables is called analysis of covariance. Although the model is more complex by including the extra variables, the profit is that the error variance is reduced.

Another very useful class of models is mixed models. Mixed models contain both fixed and random effects. They are particularly useful in social sciences when we have repeated measurements. Moreover, in the case of missing data, which are very common in sample surveys, mixed models offer a strong alternative to methods like ANOVA for repeated measures. Their drawback is that estimation is more difficult along with the fact that we end up to have a more complex model.

A useful class of models is also the semiparametric models (or even better the semiparametric regression models). These regression models include both parametric and nonparametric components. They are used when the usual parametric models do not have a satisfactory performance. More about nonparametric methods are given in Section 5.

Another very useful method is robust regression. Keeping in mind the usefulness of linear regression, its wide applicability and acceptance between the researchers it is natural to propose a method that overcomes the difficulty to fulfill its assumptions. Robust regression is used to avoid the effect of outliers. One approach is to use the M-estimators and another one is to replace the normal distribution in the assumptions with a heavy-tailed distribution.

Undoubtedly methods like linear regression and ANOVA have been used to an enormous extent in social sciences but many times without the proper accuracy in the details. We believe that much of the work done could be improved using the more advanced models presented in this section. For more details the reader could refer to Christensen (2011) and Rencher and Schaalje (2008) . For robust regression a useful reference is Rousseeuw and Leroy (1987) .

5. Nonparametric methods

In social statistics the vast amount of research is based on parametric methods. However, many parametric methods are based on strong assumptions that are disregarded most of the times. This has serious effect on the justification of the results.

The alternative in this case is to use nonparametric statistical methods. Nonparametric statistics do not rely on a specific family of probability distributions and there is no assumption about the probability distributions of the variables used. Therefore it is an ideal collection of methods for handling real data that most of the times fail to follow these strong assumptions of parametric inference.

There is a number of techniques that are already popular among the researchers in social sciences. Such techniques are certain hypothesis tests like Wilcoxon Signed rank test, Mann-Whitney test and Kruskal-Wallis tests. Other used techniques are the Spearman correlation coefficient, the runs test and normality tests. For a detailed review of such techniques the interested reader can refer to Corder and Foreman (2009) .

However, there is a number of other nonparametric methods that have been developed and are already famous among statisticians that have not gained much attention between the researchers in social sciences. Such methods are the jackknife and the bootstrap methods. Jacknife can be used to compute the bias and the variance of an estimator whereas bootstrap estimates the variance and the distribution of a statistic or it is used to construct confidence intervals. It must be noted that both these methods are computationally demanding. Nevertheless, they can be very useful in social sciences especially in the cases of complex estimators of parameters that need to be further studied.

Another useful method is nonparametric regression. The usual linear regression is a heavily used method in social sciences. However, its assumptions are very rarely referred due to the fact that they rarely hold. Nonparametric regression is a solution in that case. It is able to answer the initially stated problem that led to regression with flexibility in terms of the assumed model. Other interesting nonparametric methods are the ones used for density estimation like cross-validation and density estimation. These methods estimate the probability distribution function using just the data. They can be used in cases where the distribution of the data is unknown and difficult to be computed analytically. If a researcher is able to compute the distribution function of the variable or variables under study then he/she can obtain statistical methodologies like confidence intervals or hypothesis testing making the decision process easier and credible. For more details about these methods the interested reader could refer to Wasserman (2006) .

6. Multivariate methods

Usually in social sciences and generally in real problems more than one variable is involved. These variables need to be considered together since most of the times they are related. Several methods have been developed for the analysis of such data. These methods include among others cluster analysis, correspondence analysis, principal component analysis and factor analysis.

One of the main goals of multivariate analysis is classification. Cluster analysis is a method of classification which aims to group individuals (objects) so that those allocated to a particular group are, in a way, considered to be close together. The data used in cluster analysis are a data matrix where the columns are used for the objects and the rows for the attributes that describe the object. The output of a cluster analysis is the clusters that are used to characterize objects as similar or not. In hierarchical cluster analysis the clusters appear as a tree (they have hierarchy). In nonhierarchical cluster analysis, the number of clusters are determined by the researcher which have to be less than the number of objects. Both of these techniques are processed through statistical software. The allocation of people in similar groups is very important for a social scientist since it gives him the ability to pin point the special characteristics of these groups.

Correspondence analysis is an exploratory technique that helps a scientist to analyze multi-way frequency tables. Its main goal is to plot the data using less dimensions to identify their key features. The data used in this method have to be nonnegative and they should appear in a data table. Correspondence analysis aims to display data tables in two-dimensional spaces, called maps. The idea behind this method is that the model must follow the data, and not the opposite. In its simplest form we have a variable that we want to model and several explanatory variables. All these variables are frequencies appearing in one or more contingency tables. We use cross-tabulation for each of the explanatory variables and the variable we want to model to identify the level of their association. A technique which is also used, is to stack the tables before the application of correspondence analysis to reveal the relationship of the variable we want to model with the explanatory variables in the same map.

Principal component analysis is used to summarize p -correlated variables by a smaller number of uncorrelated variables. These variables contain most of the information that exist in the original set of variables. Keeping in mind the vast amount of data a social scientist has at hand today, we may conjecture that this technique is very important. The fact that we end up with a smaller number of variables, demands less computational power to perform the analysis of the remaining variables. Moreover, the fact that the variables are uncorrelated makes the analysis easier since the techniques used do not have to consider a relationship between the variables used. However, there a number of drawbacks. First of all, the fact that a piece of information is lost may affect the conclusions of the analysis. Moreover, if we begin with thousands of variables (which is not rare today) we may have to work with a lot of variables even after the application of principal component analysis to retain most of the information in the data.

As we already stated in the introduction sometimes in social science research, we cannot measure the variable or variables that we are interested in a direct way. These variables are called latent variables and commonly they are called factors. An example of a latent variable is human intelligence. In Factor analysis we try to relate the observable to the unobservable variables by a probability model to make statistical inference. The main objective of the analysis is to select the number of the latent variables that have to be used to explain the correlations between the unobservable variable and to interpret them. Another objective is to predict the values of the latent variables that produced the observable variables. In factor analysis the researcher regresses each of the observed variables on the set of the latent variables. Usually after the computation of the factors a social scientist tries to “name” them based on the numerical findings. However, since there is not a specific way to perform this action, the result of this step is sometimes not properly elaborated. For all the above methods indicative references are Everitt (1993) , Greenacre (2007) , Jolliffe (2002) and Bartholomew and Knott (2011) .

Apart from these well-known methods there are also some other methods equally important but less used. These methods are path analysis, structural equation modeling and multilevel modeling.

Path analysis is concerned with causation. Specifically it uses regression methods to identify patterns of causation in networks. In the beginning path analysis starts with a network of variables to specify the paths of causation. Usually, a cause and effect relationship assumes that there are a number of relationships and some variables that are believed to be caused by others, appear to affect other variables. A regression model cannot identify such a case because it can merely use one dependent variable. In path analysis all the necessary regression models considered, account for all the relationships needed.

Structural theory tries to give the structural relationships between constructs. This theory is represented by a structural model using a number of equations. These equations are usually accompanied by a proper diagram indicating the relationships. In other words, structural equation modeling is a method that tries to estimate the relationship between latent variables. This relationship can be linear or non-linear. The advantage of this method is that it allows us to test hypotheses on the relationships between observed variables and latent variables and also between the latent variables themselves.

Multilevel modeling is used to analyze data involving clusters. Specifically, in social research we are often concerned with the relationship between individuals and the groups they belong. This relationship actually leads to nested data, that is individuals nested within groups. For example in education students are nested within schools. The performance of a student in a series of exams could be affected by both characteristics of the student and of the school he/she attended.

For path analysis, structural equation modeling and multilevel modeling, the interested reader can refer to Agresti and Finlay (2009) , Bartholomew et al. (2008) and Timm (2002) .

7. Categorical data

Usually in social sciences researchers have to analyze categorical data. A categorical variable can take a limited number of specific discrete values. Usually such values occur for example when respondents are assigned in groups or when a property holds or not. In social sciences the different categories of a categorical variable often measures attitudes and opinions.

Categorical variables with a natural ordering are called ordinal variables. Categorical variables without ordering are called nominal variables. Methods designed for ordinal variables cannot be used with nominal variables due to the fact that nominal variables do not have ordered categories. Methods designed for nominal variables can be used with nominal or ordinal variables, since they only require a categorical scale.

The most famous models for analyzing categorical data are logistic regression models. Logistic regression can be used with continuous and discrete predictors ( Agresti (2007) ). Loglinear models are used to analyze associations among multiple categorical response variables. A log-linear model can be transformed using logarithm to a polynomial function of the parameters of the model. This is very helpful since the researcher can use linear regression ( Azen and Walker (2011) ).

A broad class of models is the generalized linear models. These models are a generalization of ordinary linear regression in the sense that it allows the distribution of the error to be different from the normal distribution. Another class of models is those that are used to analyze repeated measures data or longitudinal data. That kind of data is repeated observations of the same variables over several periods of time. One feature that must be taken into consideration is that data are correlated since the same subjects are measured over time ( Lawal (2003) ).

We may say that in general researchers in social sciences could rely more on the abovementioned models for the analysis of categorical data. These models are not very popular among researchers who tend to rely more on descriptive measures. We believe that the practitioners in the area could benefit a lot from the already developed methods.

8. Time series

Time series is a sequence of observations on a variable of interest with chronological order. That kind of data is quite natural in some of the fields in social sciences like economics. The observations in a time series are considered dependent. Time series analysis is a collection of techniques for the analysis of this time dependence.

There are a lot of different approaches to handle time series data. A first approach is to use the autoregressive models or the moving average models. The autoregressive model (AR) assumes that there is linear dependence of the variable we study with its own previous values. The moving average (MA) model is a linear regression of the current value of the series against current and previous (in terms of time) error terms.

Another class of models are the autoregressive moving average (ARMA) models. We use the notation ARMA( p , q ) to define a model with p autoregressive terms and q moving-average terms. A generalization of this model is the autoregressive integrated moving average (ARIMA) model. This model is generally referred as ARIMA( p , d, q ) where parameters p , d, and q are non-negative integers that refer to the order of the autoregressive, integrated and moving average parts of the model respectively. All the above mentioned models (AR, MA, ARMA, ARIMA) form among other techniques the Box-Jenkins method for modeling time series. For more details Box et al. (2008) .

Another class of time series models, especially useful in econometrics, are the autoregressive conditional heteroskedasticity (ARCH) models. In ARCH models we assume that the variance of the current error term is a function of the actual sizes of the previous time periods’ error terms. ARCH models have been extensively used to model financial time series. A generalization of the ARCH models is the generalized autoregressive conditional heteroskedasticity models (GARCH). In GARCH models we assume that the error variance is modeled by an ARMA model. There are a number of newer model proposals based on ARCH and GARCH models. The interested reader about ARCH models can refer to Xekalaki and Degiannakis (2010) .

Another interesting characteristic in time series is forecasts. Apparently, it has attracted the interest of researchers in various fields. Several techniques on this very interesting issue have been proposed. Methods and examples of applications are given in Bisgaard and Kulahci (2011) .

The research in social sciences, using the already stated models for time series, mainly appears in economics and marketing. We strongly believe that researchers in other areas of social sciences could benefit from these models also.

9. Data mining

Data mining is a collection of techniques used to find patterns in a set of data. They are extremely important in the analysis of large data sets of social phenomena. Other names that refer to the same collection of techniques are machine learning and predictive analytics. During the last years there is an increasing interest in these techniques although most of them are known for decades. We have to note here that the use of a computer is compulsory to run these techniques and moreover that if we have large data sets the larger the amount of data the more computational power we need.

The computational methods that comprise the field of data mining derive from the areas of statistics and artificial intelligence. These techniques are used to find meaningful associations between related variables usually between a large number of variables. These structures help the practitioner to draw useful conclusions about his/her research questions.

An important feature that is one of the objectives of a data mining analysis is the generalization of the results. To be more specific, if after an analysis of the data at hand using data mining techniques we conclude that there are some important patterns, then we would also like to find that these patterns exist and in the data that we will gather in future. This generalization is very important for drawing conclusions that are irrespective of the collected data the specific time we run the analysis.

If we consider the predictive dimension of data mining we can refer to the two important conclusions of such an analysis. The first conclusion is that after we reach a useful and meaningful model we can use it to predict the variable under study using some or all the remaining variables. Obviously, such a conclusion gives the researcher the ability to compute the values of the dependent variable given the values of the independent variables. The second conclusion is that the researcher is able to comment about the relationship between the dependent and the independent variables.

Another characteristic we need to highlight is the need to know as much as possible about the data and the process. The definition of the variables, the way they are measured and their interrelation in terms of the case studied are extremely important to the researcher to assist him reach a meaningful conclusion. Additionally, since the data are most of the times in vast numbers there is the need to store, process and compute them. Therefore, it is highly probable that knowledge of databases and parallel computing will be compulsory for the application of data mining techniques.

Keeping in mind the vast amount of social data that are gathered in today’s world using classical ways (e.g. questionnaires) along with the use of mobile technologies, social networks, texts, photographs, videos and all the different types of human activities that are transformed to data we can easily conclude that it is not a rare event to have to analyze thousands of variables with many cases in each of them. In such cases we can say that we end up with big data (data with high volume, high velocity and high variety). This fact highlights the need to use data mining techniques that can handle such amount of data. More information and detailed representation of data mining techniques can be found in Hastie et al. (2009) and Azzalini and Scarpa (2010) .

10. Conclusions

In this paper, we reviewed some statistical methods useful in the area of social sciences. Sampling techniques, regression analysis, analysis of variance, parametric and nonparametric models along with multivariate methods were presented. Categorical data analysis techniques, time series methods and data mining were also presented. Indicative references in all of these areas are also given.

Statistical methods have played a very important role in social sciences. In every applied research effort statistical techniques are compulsory to reach a non-questionable conclusion. We strongly believe that advanced statistical methods can be employed heavily in this area. It seems that researchers rely more on classical statistical methods although they could benefit from the use of newer and advanced techniques.

Agresti , A. ( 2007 ), An Introduction to Categorical Data Analysis , John Wiley , New York, NY .

Agresti , A. and Finlay , B. ( 2009 ), Statistical Methods for the Social Sciences , 4th ed. , Pearson/Prentice Hall , NJ .

Azen , R. and Walker , C.M. ( 2011 ), Categorical Data Analysis for the Behavioral and Social Sciences , Routledge , New York, NY .

Azzalini , A. and Scarpa , B. ( 2010 ), Data Analysis and Data Mining , Oxford University Press , New York, NY .

Bartholomew , D.J. and Knott , M. ( 2011 ), Latent Variable Models and Factor Analysis , 2nd ed. , Vol. 7 , Kendall’s Library of Statistics , Arnold .

Bartholomew , D.J. , Steele , F. , Moustaki , I. and Galbraith , J.I. ( 2008 ), Analysis of Multivariate Social Science Data , 2nd ed. , CRC Press , New York, NY .

Bisgaard , S. and Kulahci , M. ( 2011 ), Time Series Analysis and Forecasting by Example , John Wiley , New York, NY .

Box , G.E.P. , Jenkins , G.M. and Reinsel , G.C. ( 2008 ), Time Series Analysis: forecasting and Control , 4th ed. , John Wiley , New York, NY .

Christensen , R. ( 2011 ), Plane Answers to Complex Questions: The Theory of Linear Models , Springer .

Cohen , J. ( 1988 ), Statistical Power for Behavioral Sciences , Lawrence Erlbaum Assoc ., Mahwah, NJ .

Cohen , B.H. and Lea , R.B. ( 2004 ), Essentials of Statistics for the Social and Behavioral Sciences , John Wiley , New York, NY .

Corder , G.W. and Foreman , D.I. ( 2009 ), Nonparametric Statistics for Non-Statisticians: A Step-by-Step Approach , John Wiley , New York, NY .

Dodge , Y. ( 2008 ), The Concise Encyclopedia of Statistics , Springer .

Everitt , B. ( 1993 ), Cluster Analysis , Arnold , London .

Greenacre , M.J. ( 2007 ), Correspondence Analysis in Practice , Chapman and Hall , Boca Raton .

Hastie , T. , Tibshirani , R. and Friedman , J. ( 2009 ), The Elements of Statistical Learning: data Mining, Inference, and Prediction , Springer , New York, NY .

Jacob , R. and Zhu , P. ( 2012 ), A Practical Guide to Regression Discontinuity , mdrc .

Jolliffe , I.T. ( 2002 ), Principal Component Analysis , 2nd ed. , Springer , New York, NY .

Joseph , L. , Burger , R.D. and Belisle , P. ( 1997 ), “ Bayesian and mixed bayesian/likelihood criteria for sample size determination ”, Statistics in Medicine , Vol. 16 No. 7 , pp. 769 - 789 .

Kish , L. ( 1995 ), Survey Sampling , John Wiley , New York, NY .

Kutner , M.H. , Nachtsheim , C.J. , Neter , J. and Li , W. ( 2005 ), Applied Linear Statistical Models , McGraw-Hill/Irwin , New York, NY .

Lawal , B.H. ( 2003 ), Categorical Data Analysis with SAS and SPSS Applications , Lawrence Erlbaum Associates , NJ .

Lenth , R.V. ( 2001 ), “ Some practical guidelines for effective sample size calculations ”, American Statistician , Vol. 55 No. 3 , pp. 187 - 193 .

Lovric , M. ( 2011 ), International Encyclopedia of Statistical Science , Springer .

Rencher , A.C. and Schaalje , G.B. ( 2008 ), Linear Models in Statistics , John Wiley , New York, NY .

Riley-Tilman , T.C. and Burns , M.K. ( 2009 ), Evaluating Educational Interventions , The Guilford Press , New York, NY .

Rousseeuw , P.J. and Leroy , A.M. ( 1987 ), Robust Regression and Outlier Detection , John Wiley , New York, NY .

Shuster , J.J. ( 1990 ), Handbook of Sample Size Guidelines for Clinical Trials , CRC Press , Boca Raton, FL .

Timm , N.H. ( 2002 ), Applied Multivariate Analysis , Springer , New York, NY .

Wasserman , L. ( 2006 ), All of Nonparametric Statistics , Springer , New York, NY .

Xekalaki , E. and Degiannakis , S. ( 2010 ), ARCH Models for Financial Applications , John Wiley , New York, NY .

Acknowledgements

This article is an invited submission and was not peer reviewed.

Corresponding author

Related articles, we’re listening — tell us what you think, something didn’t work….

Report bugs here

All feedback is valuable

Please share your general feedback

Join us on our journey

Platform update page.

Visit emeraldpublishing.com/platformupdate to discover the latest news and updates

Questions & More Information

Answers to the most commonly asked questions here

Want to create or adapt books like this? Learn more about how Pressbooks supports open publishing practices.

2.2 Paradigms, theories, and how they shape a researcher’s approach

Learning objectives.

• Define paradigm, and describe the significance of paradigms
• Identify and describe the four predominant paradigms found in the social sciences
• Define theory
• Describe the role that theory plays in social work research

The terms paradigm and theory are often used interchangeably in social science, although social scientists do not always agree whether these are identical or distinct concepts. This text makes a clear distinction between the two ideas because thinking about each concept as analytically distinct provides a useful framework for understanding the connections between research methods and social scientific ways of thinking.

Paradigms in social science

  For our purposes, we’ll define paradigm as a way of viewing the world (or “analytic lens” akin to a set of glasses) and a framework from which to understand the human experience (Kuhn, 1962). It can be difficult to fully grasp the idea of paradigmatic assumptions because we are very ingrained in our own, personal everyday way of thinking. For example, let’s look at people’s views on abortion. To some, abortion is a medical procedure that should be undertaken at the discretion of each individual woman. To others, abortion is murder and members of society should collectively have the right to decide when, if at all, abortion should be undertaken. Chances are, if you have an opinion about this topic, you are pretty certain about the veracity of your perspective. Then again, the person who sits next to you in class may have a very different opinion and yet be equally confident about the truth of their perspective. Who is correct?

You are each operating under a set of assumptions about the way the world does—or at least should—work. Perhaps your assumptions come from your political perspective, which helps shape your view on a variety of social issues, or perhaps your assumptions are based on what you learned from your parents or in church. In any case, there is a paradigm that shapes your stance on the issue. Those paradigms are a set of assumptions. Your classmate might assume that life begins at conception and the fetus’ life should be at the center of moral analysis. Conversely, you may assume that life begins when the fetus is viable outside the womb and that a mother’s choice is more important than a fetus’s life. There is no way to scientifically test when life begins, whose interests are more important, or the value of choice. They are merely philosophical assumptions or beliefs. Thus, a pro-life paradigm may rest in part on a belief in divine morality and fetal rights. A pro-choice paradigm may rest on a mother’s self-determination and a belief that the positive consequences of abortion outweigh the negative ones. These beliefs and assumptions influence how we think about any aspect of the issue.

In Chapter 1, we discussed the various ways that we know what we know. Paradigms are a way of framing what we know, what we can know, and how we can know it. In social science, there are several predominant paradigms, each with its own unique ontological and epistemological perspective. Recall that ontology is the study of what is real, and epistemology is the study of how we come to know what is real. Let’s look at four of the most common social scientific paradigms that might guide you as you begin to think about conducting research.

The first paradigm we’ll consider, called positivism, is the framework that likely comes to mind for many of you when you think of science. Positivism is guided by the principles of objectivity, knowability, and deductive logic. Deductive logic is discussed in more detail in next section of this chapter. The positivist framework operates from the assumption that society can and should be studied empirically and scientifically. Positivism also calls for a value-free science, one in which researchers aim to abandon their biases and values in a quest for objective, empirical, and knowable truth.

Another predominant paradigm in social work is social constructionism . Peter Berger and Thomas Luckman (1966) are credited by many for having developed this perspective in sociology. While positivists seek “the truth,” the social constructionist framework posits that “truth” varies. Truth is different based on who you ask, and people change their definitions of truth all the time based on their interactions with other people. This is because we, according to this paradigm, create reality ourselves (as opposed to it simply existing and us working to discover it) through our interactions and our interpretations of those interactions. Key to the social constructionist perspective is the idea that social context and interaction frame our realities.

Researchers operating within this framework take keen interest in how people come to socially agree, or disagree, about what is real and true. Consideration of how meanings of different hand gestures vary across different regions of the world aptly demonstrates that meanings are constructed socially and collectively. Think about what it means to you when you see a person raise their middle finger. In the United States, people probably understand that person isn’t very happy (nor is the person to whom the finger is being directed). In some societies, it is another gesture, such as the thumbs up gesture, that raises eyebrows. While the thumbs up gesture may have a particular meaning in North American culture, that meaning is not shared across cultures (Wong, 2007). So, what is the “truth” of the middle finger or thumbs up? It depends on what the person giving it intended, how the person receiving it interpreted it, and the social context in which the action occurred.

It would be a mistake to think of the social constructionist perspective as only individualistic. While individuals may construct their own realities, groups—from a small one such as a married couple to large ones such as nations—often agree on notions of what is true and what “is.” In other words, the meanings that we construct have power beyond the individual people who create them. Therefore, the ways that people and communities work to create and change such meanings is of as much interest to social constructionists as how they were created in the first place.

A third paradigm is the critical paradigm. At its core, the critical paradigm is focused on power, inequality, and social change. Although some rather diverse perspectives are included here, the critical paradigm, in general, includes ideas developed by early social theorists, such as Max Horkheimer (Calhoun, Gerteis, Moody, Pfaff, & Virk, 2007), and later works developed by feminist scholars, such as Nancy Fraser (1989). Unlike the positivist paradigm, the critical paradigm posits that social science can never be truly objective or value-free. Further, this paradigm operates from the perspective that scientific investigation should be conducted with the express goal of social change in mind. Researchers in the critical paradigm might start with the knowledge that systems are biased against, for example, women or ethnic minorities. Moreover, their research projects are designed not only to collect data, but also change the participants in the research as well as the systems being studied. The critical paradigm not only studies power imbalances but seeks to change those power imbalances.

Finally, postmodernism is a paradigm that challenges almost every way of knowing that many social scientists take for granted (Best & Kellner, 1991). While positivists claim that there is an objective, knowable truth, postmodernists would say that there is not. While social constructionists may argue that truth is in the eye of the beholder (or in the eye of the group that agrees on it), postmodernists may claim that we can never really know such truth because, in the studying and reporting of others’ truths, the researcher stamps their own truth on the investigation. Finally, while the critical paradigm may argue that power, inequality, and change shape reality and truth, a postmodernist may in turn ask whose power, whose inequality, whose change, whose reality, and whose truth. As you might imagine, the postmodernist paradigm poses quite a challenge for researchers. How do you study something that may or may not be real or that is only real in your current and unique experience of it? This fascinating question is worth pondering as you begin to think about conducting your own research. Part of the value of the postmodern paradigm is its emphasis on the limitations of human knowledge. Table 2.1 summarizes each of the paradigms discussed here.

Let’s work through an example. If we are examining a problem like substance abuse, what would a social scientific investigation look like in each paradigm? A positivist study may focus on precisely measuring substance abuse and finding out the key causes of substance abuse during adolescence. Forgoing the objectivity of precisely measuring substance abuse, social constructionist study might focus on how people who abuse substances understand their lives and relationships with various drugs of abuse. In so doing, it seeks out the subjective truth of each participant in the study. A study from the critical paradigm would investigate how people who have substance abuse problems are an oppressed group in society and seek to liberate them from external sources of oppression, like punitive drug laws, and internal sources of oppression, like internalized fear and shame. A postmodern study may involve one person’s self-reported journey into substance abuse and changes that occurred in their self-perception that accompanied their transition from recreational to problematic drug use. These examples should illustrate how one topic can be investigated across each paradigm.

Social science theories

Much like paradigms, theories provide a way of looking at the world and of understanding human interaction. Paradigms are grounded in big assumptions about the world—what is real, how do we create knowledge—whereas theories describe more specific phenomena. A common definition for theory in social work is “a systematic set of interrelated statements intended to explain some aspect of social life” (Rubin & Babbie, 2017, p. 615). At their core, theories can be used to provide explanations of any number or variety of phenomena. They help us answer the “why” questions we often have about the patterns we observe in social life. Theories also often help us answer our “how” questions. While paradigms may point us in a particular direction with respect to our “why” questions, theories more specifically map out the explanation, or the “how,” behind the “why.”

Introductory social work textbooks introduce students to the major theories in social work—conflict theory, symbolic interactionism, social exchange theory, and systems theory. As social workers study longer, they are introduced to more specific theories in their area of focus, as well as perspectives and models (e.g., the strengths perspective), which provide more practice-focused approaches to understanding social work.

As you may recall from a class on social work theory, systems theorists view all parts of society as interconnected and focus on the relationships, boundaries, and flows of energy between these systems and subsystems (Schriver, 2011). Conflict theorists are interested in questions of power and who wins and who loses based on the way that society is organized. Symbolic interactionists focus on how meaning is created and negotiated through meaningful (i.e., symbolic) interactions. Finally, social exchange theorists examine how human beings base their behavior on a rational calculation of rewards and costs.

Just as researchers might examine the same topic from different levels of inquiry or paradigms, they could also investigate the same topic from different theoretical perspectives. In this case, even their research questions could be the same, but the way they make sense of whatever phenomenon it is they are investigating will be shaped in large part by theory. Table 2.2 summarizes the major points of focus for four major theories and outlines how a researcher might approach the study of the same topic, in this case the study of substance abuse, from each of the perspectives.

Within each area of specialization in social work, there are many other theories that aim to explain more specific types of interactions. For example, within the study of sexual harassment, different theories posit different explanations for why harassment occurs. One theory, first developed by criminologists, is called routine activities theory. It posits that sexual harassment is most likely to occur when a workplace lacks unified groups and when potentially vulnerable targets and motivated offenders are both present (DeCoster, Estes, & Mueller, 1999). Other theories of sexual harassment, called relational theories, suggest that a person’s relationships, such as their marriages or friendships, are the key to understanding why and how workplace sexual harassment occurs and how people will respond to it when it does occur (Morgan, 1999). Relational theories focus on the power that different social relationships provide (e.g., married people who have supportive partners at home might be more likely than those who lack support at home to report sexual harassment when it occurs). Finally, feminist theories of sexual harassment take a different stance. These theories posit that the way our current gender system is organized, where those who are the most masculine have the most power, best explains why and how workplace sexual harassment occurs (MacKinnon, 1979). As you might imagine, which theory a researcher applies to examine the topic of sexual harassment will shape the questions the researcher asks about harassment. It will also shape the explanations the researcher provides for why harassment occurs.

For an undergraduate student beginning their study of a new topic, it may be intimidating to learn that there are so many theories beyond what you’ve learned in your theory classes. What’s worse is that there is no central database of different theories on your topic. However, as you review the literature in your topic area, you will learn more about the theories that scientists have created to explain how your topic works in the real world. In addition to peer-reviewed journal articles, another good source of theories is a book about your topic. Books often contain works of theoretical and philosophical importance that are beyond the scope of an academic journal.

Paradigm and theory in social work

Theories, paradigms, levels of analysis, and the order in which one proceeds in the research process all play an important role in shaping what we ask about the social world, how we ask it, and in some cases, even what we are likely to find. A micro-level study of gangs will look much different than a macro-level study of gangs. In some cases, you could apply multiple levels of analysis to your investigation, but doing so isn’t always practical or feasible. Therefore, understanding the different levels of analysis and being aware of which level you happen to be employing is crucial. One’s theoretical perspective will also shape a study. In particular, the theory invoked will likely shape not only the way a question about a topic is asked but also which topic gets investigated in the first place. Further, if you find yourself especially committed to one theory over another, it may limit the kinds of questions you pose. As a result, you may miss other possible explanations.

The limitations of paradigms and theories do not mean that social science is fundamentally biased. At the same time, we can never claim to be entirely value free. Social constructionists and postmodernists might point out that bias is always a part of research to at least some degree. Our job as researchers is to recognize and address our biases as part of the research process, if an imperfect part. We all use our own approaches, be they theories, levels of analysis, or temporal processes, to frame and conduct our work. Understanding those frames and approaches is crucial not only for successfully embarking upon and completing any research-based investigation, but also for responsibly reading and understanding others’ work.

Spotlight on UTA School of Social Work

Catherine labrenz connects social theory and child welfare research.

When Catherine LaBrenz, an assistant professor at the University of Texas at Arlington’s School of Social Work was a child welfare practitioner, she noticed that several children who had reunified with their biological parents from the foster care system were re-entering care because of continued exposure to child maltreatment. As she observed the challenging behaviors these children often presented, she wondered how the agency might better support families to prevent children from re-entering foster care after permanence. In her doctoral studies, she used her practice experience to form a research project with the goal of better understanding how agencies could better support families post-reunification.

From a critical paradigm, Dr. LaBrenz approached this question with the understanding that families that come into contact with child welfare systems often experience disadvantage and are subjected to unequal power distributions when accessing services, going to court, and participating in case decision-making (LaBrenz & Fong, 2016). Furthermore, the goal of this research was to change some of the aspects of the child welfare system, particularly within the practitioner’s agency, to better support families.

To better understand why some families may be more at-risk for multiple entries into foster care, Dr. LaBrenz began with an extensive literature review that identified diverse theories that explained factors at the child, family, and system- level that could impact post-permanence success. Figure 2.1 displays the micro-, meso-, and macro-level theories that she and her research team identified and decided to explore further.

At the child-level, Attachment theory posits that consistent, stable nurturing during infancy impacts children’s ability to form relationships with others throughout their life (Ainsworth, Blehar, Waters, & Wall, 1978; Bowlby, 1969). At the family-level, Family systems theory posits that family interactions impact functioning among all members of a family unit (Broderick 1971). At the macro-level, Critical race theory (Delgado & Stefancic, 2001) can help understand racial disparities in child welfare systems. Moreover, Systems theory (Bronfenbrenner, 1986) can help examine interactions among the micro-, meso- and macro-levels to assess diverse systems that impact families involved in child welfare services.

In the next step of the project, national datasets were used to examine child-, family-, and system- factors that impacted rates of successful reunification, or reunification with no future re-entries into foster care. Then, a systematic review of the literature was conducted to determine what evidence existed for interventions to increase rates of successful reunification. Finally, a different national dataset was used to examine how effective diverse interventions were for specific groups of families, such as those with infants and toddlers.

Figure 2.2 displays the principal findings from the research project and connects each main finding to one of the theoretical frameworks.

The first part of the research project found parents who felt unable to cope with their parental role, and families with previous attachment disruptions, to have higher rates of re-entry into foster care. This connects with Attachment theory, in that families with more instability and inconsistency in caregiving felt less able to fulfill their parental roles, which in turn led to further disruption in the child’s attachment.

With regards to family-level theories, Dr. LaBrenz found that family-level risk and protective factors were more predictive of re-entry to foster care than child- or agency-level factors. The systematic review also found that interventions that targeted parents, such as Family Drug Treatment Courts, led to better outcomes for children and families. This aligns with Family systems theory in that family-centered interventions and targeting the entire family leads to better family functioning and fewer re-entries into foster care.

In parallel, the systematic review concluded that interventions that integrated multiple systems, such as child welfare and substance use, increased the likelihood of successful reunification. This supports Systems theory, in that multiple systems can be engaged to provide ongoing support for families in child welfare systems (Trucco, 2012). Furthermore, the results from the analyses of the national datasets found that rates of re-entry into foster care for African American and Latino families varied significantly by state. Thus, racial and ethnic disparities remained in some, but not all, state child welfare systems.

Overall, the findings from the research project supported Attachment theory, Family systems theory, Systems theory, and Critical race theory as guiding explanations for why some children and families experience foster care re-entry while others do not. Dr. LaBrenz was able to present these findings and connect them to direct implications for practices and policies that could support attachment, multi-system collaborations, and family-centered practices.

Key Takeaways

• Paradigms shape our everyday view of the world.
• Researchers use theory to help frame their research questions and to help them make sense of the answers to those questions.
• Applying the four key theories of social work is a good start, but you will likely have to look for more specific theories about your topic.
• Critical paradigm- a paradigm in social science research focused on power, inequality, and social change
• Paradigm- a way of viewing the world and a framework from which to understand the human experience
• Positivism- a paradigm guided by the principles of objectivity, knowability, and deductive logic
• Postmodernism- a paradigm focused on the historical and contextual embeddedness of scientific knowledge and a skepticism towards certainty and grand explanations in social science
• Social constructionism- a paradigm based on the idea that social context and interaction frame our realities
• Theory- “a systematic set of interrelated statements intended to explain some aspect of social life” (Rubin & Babbie, 2017, p. 615)

Image attributions

point mold and cloud mold by tasaikensuke CC-0

why by GDJ CC-0

Foundations of Social Work Research Copyright © 2020 by Rebecca L. Mauldin is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License , except where otherwise noted.

Share This Book

Limitations of the study

Link to guide  here.

Who It’s For: New and seasoned researchers who are unfamiliar with the different possible limitations of their study and/or are unsure of how to discuss limitations in their research paper.

Why We Love It: This resource is easy-to-follow and is great to start with and return to as you consider and write about the limitations of your study. It not only explains possible limitations of your research, but offers a hand ful of useful tips for writing about these limitations effectively. 

Improving Care for Older Adults

Photo Report from the Population Aging Research Center 2024 Retreat

Defining the health and health care problems and prospects of aging americans.

• Hoag Levins
• Share this page on Twitter
• Share this page on Facebook
• Share this page on LinkedIn

The University of Pennsylvania Population Aging Research Center’s (PARC) 2024 Retreat on May 3 brought together an interdisciplinary group of academics whose work exemplifies the organization’s reputation as an international leader in research on the structure, organization, health, and well-being of aging human populations. The event’s 22 podium presentations and session posters covered the gamut of health issues ranging from accelerated aging among formerly incarcerated older adults and the financial penalties borne by unpaid family caregivers, to the lack of sufficient infrastructure to meet minimal long-term care needs of older adults and the racial life expectancy disparities that speak of the structural racism that has defined the health outlook for so many Americans. Here’s a look at some of what happened in the meeting at the Penn School of Arts and Sciences’ McNeil Building:

Editor, Digital Publications

More ldi news.

In Their Own Words

Health Care Access & Coverage

Bipartisan Improvements May Be Coming For One of the Biggest Problems in Health Care

Dually Eligible Americans on Medicare and Medicaid Are Getting the Attention They Deserve

• Rachel M. Werner, MD, PhD

The Population Health Calamity Evidenced in U.S. Life Expectancy Statistics

And What it Portends for Aging Adult Health in the Coming Decades

Health Equity

Parents Want Providers to Log Their Children’s Social Needs but with Transparency and Respect

Parents Also Fear Disclosing Social Needs Will Lead to Child Welfare Referrals, a Study by LDI Fellows Finds

• Miles Meline, MBE

News | Video

Penn Study in Navajo Nation Boosts Guideline Heart Failure Drug Uptake by 53%

LDI Senior Fellow Lauren Eberly Leads Team in New Mexico

Growing Health Worker Migration to the U.S. and U.K. Raises Fairness and Training Issues

The U.S. Needs to Help Fix Migration’s Root Causes and Build Better Training Systems

• Farouk Dako, MD, MPH

Health Care Algorithms Can Improve or Worsen Racial and Ethnic Disparities

Users Need To Understand Existing Disparities and Why Race Can Be Imprecise

Exploring the Policies That Are Closing Rural Hospitals

A Penn LDI Virtual Seminar Looks at Trends Driving the Crisis

Population Health

4 Ways Health Officials Can Communicate Better in the Next Pandemic

COVID-19’s Key Challenges Provide Insight for Future Outbreaks