Can Generative AI Suggest Relevant Theories for a Research Problem?

Can AI Find a Theory That Actually Fits Your Research Problem?

You have a research problem but are uncertain which theory could explain it. You describe your study to a generative AI tool and ask for theoretical recommendations. Within seconds, it suggests several established theories, identifies their proponents, and explains why each might be relevant.

The recommendations sound reasonable. Some theories are familiar, while others are entirely new to you. But how do you determine whether AI has identified a genuinely appropriate theoretical perspective or simply matched your research keywords with frequently discussed theories?

Selecting a theory affects how you explain your research problem, interpret relevant concepts, and justify your investigation. A convincing AI recommendation is therefore a starting point for theoretical inquiry, not a sufficient basis for choosing a framework.

AI Can Suggest Candidate Theories, but Relevance Must Be Established

How to Evaluate Theories Suggested by Generative AI

How Does Generative AI Suggest Research Theories?

When you describe a research problem, a generative AI system can identify patterns connecting your description with theoretical terminology encountered during training. Depending on the system, it may also retrieve relevant documents or search external scholarly information.

For example, a question about students' intentions to adopt educational technology might produce recommendations involving the Technology Acceptance Model or the Unified Theory of Acceptance and Use of Technology. These are established theoretical models associated with technology acceptance and use.

However, the fact that a theory frequently appears alongside certain concepts does not establish that it can adequately explain your particular research problem.

AI may produce theoretically plausible suggestions without fully accounting for the theory's underlying assumptions, original explanatory purpose, or contextual limitations.

What Makes a Theory Relevant to a Research Problem?

Theoretical relevance is primarily a question of explanatory fit. A theory should help explain the phenomenon being investigated rather than merely contain terminology resembling the research title.

Whetten (1989) identifies important elements of theoretical reasoning, including the factors involved, their relationships, the explanation for those relationships, and the conditions under which an explanation applies. These provide useful questions for evaluating AI recommendations.

These criteria are not a universal scoring instrument. They are practical questions for evaluating whether an AI recommendation deserves further consideration.

Why Might AI Recommend the Most Familiar Theory Rather Than the Best-Fitting One?

Some theories are repeatedly discussed alongside particular research topics. Generative AI may therefore recommend familiar theoretical perspectives even when a less commonly cited explanation would be more appropriate.

Consider a study examining teachers' experiences of institutional pressure when adopting artificial intelligence. A model of individual technology acceptance might explain some aspects of teachers' intentions, but it may not adequately address institutional expectations, professional autonomy, or organizational constraints.

The appropriate theory depends on what the study is actually trying to explain.

This is why researchers should examine whether AI recommendations reflect keyword associations rather than theoretical fit.

Should You Give AI Your Research Title or Your Complete Research Problem?

A research title may provide insufficient information for meaningful theoretical suggestions. Titles often omit the precise phenomenon, research context, methodological orientation, and explanatory questions.

Compare these two requests:

The second request gives AI more information about the phenomenon requiring explanation. It may produce more focused suggestions, although additional context does not guarantee accuracy.

Can AI Suggest Theories From Different Disciplines?

Yes. AI may identify perspectives from psychology, sociology, education, information systems, or organizational studies that a researcher might not initially consider.

This can be useful for interdisciplinary research. A study of digital learning, for example, may involve psychological explanations of motivation alongside sociological explanations of institutional practices.

However, transferring a theory across disciplinary contexts requires justification. A theory developed to explain individual behavior should not automatically be treated as an explanation of organizational processes.

Researchers should consider whether the theory's concepts and assumptions remain meaningful in the new context and whether previous scholarship supports the proposed application.

Should AI Rank Theories or Identify the Single Best Theory?

AI can compare candidate theories against explicitly defined criteria. It may even produce a ranking, but that ranking should not be mistaken for an independently validated assessment of theoretical suitability.

There is rarely a universally best theory for a research topic. Different theories may explain different aspects of the same phenomenon, and their suitability depends on the research questions and explanatory objectives.

Instead of asking AI to identify the best theory without qualification, ask it to explain what each theory can and cannot account for.

A structured comparison may help, but researchers should still evaluate competing theoretical perspectives using scholarly evidence.

How Can You Verify an AI-Suggested Theory?

Verification should establish more than the existence of the theory. Researchers also need to determine whether AI has accurately described its central propositions and whether those propositions apply to the proposed study.

Researchers should check AI-suggested theories against primary theoretical sources before incorporating them into a proposal or manuscript.

Can AI Invent a Theory That Does Not Exist?

Yes. Generative AI can produce nonexistent theory names, combine elements from unrelated theories, or incorrectly attribute theoretical ideas to particular scholars.

Even when a theory exists, AI may misrepresent its original scope or introduce constructs that were not part of the author's formulation.

Choosing a Theory for Students' Adoption of an AI Learning Tool

Rather than selecting the first recommendation, the researcher examines the explanatory focus of each candidate.

The researcher then consults Davis (1989), Venkatesh and colleagues (2003), and Bandura's original theoretical work.

If the research questions specifically concern students' perceived usefulness and ease of use as explanations of their intention to adopt the tool, the Technology Acceptance Model may provide an appropriate foundation.

If the study instead examines institutional influence and facilitating conditions alongside individual expectations, the researcher may find the Unified Theory of Acceptance and Use of Technology more suitable.

The decision: The researcher selects a theoretical perspective based on the explanatory requirements of the study, not on the order in which AI recommended the theories.

This does not establish that one model is universally superior. It demonstrates how AI-generated candidates can be evaluated through their theoretical mechanisms and contextual relevance.

Mistakes That Make AI-Suggested Theories Difficult to Defend

Use AI to Expand Your Options, Then Narrow Them Through Theory

AI-assisted theory exploration is most useful when you provide a well-defined research problem and request explanations of theoretical fit rather than a list of famous theories.

Consider asking AI to identify why each candidate may be unsuitable. This encourages attention to theoretical limitations that might otherwise disappear beneath a polished recommendation.

A More Useful Prompt for Theory Exploration

After evaluating the recommendations, you can begin developing a theoretically justified framework around the perspective that best addresses the research problem.

The objective is not to obtain a theory that AI approves. It is to identify a theory whose relevance you can explain and defend through scholarship.

Before Selecting an AI-Suggested Theory

Frequently Asked Questions About AI Theory Recommendations

A Suggested Theory Is a Candidate, Not a Conclusion

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.