Selecting and Applying Theories in Medical Education Research

"Grounded Theory" is a bit of a misnomer. In essence, it's a robust methodology dedicated to the collection and analysis of qualitative data. The hallmark of Grounded Theory research is its objective: to forge a new theory.

So, rather than starting with a theoretical framework to guide the research design and analysis, the process culminates in the emergence of a unique theory. Researchers embark on this journey without preconceived notions, allowing the data to guide them toward the development of an original framework.

This newly minted theory then becomes a tool for further exploration, offering a foundation for testing, refinement, and application.

Research that aims to investigate relatively unknown or less understood phenomena may not utilize theory. Methods commonly applied in these studies include qualitative interviews, focus groups, observational work, document review, Delphi analyses, and more.

While these approaches don’t always need theory, experts advise that these approaches are often enhanced by theory, leading to higher quality and more rigorous output.

When the primary aim of the research is to describe a specific situation, event, or condition without attempting to infer relationships or causality, theory may not be necessary. For example, a demographic study looking at the student body composition of various medical schools may not require theoretical framing, at least during the design process.

Theory can be brought in later to make sense of the results in the analysis stage, an approach that is discussed in greater detail below.

Research focusing on the development or refinement of techniques, tools, or methods may not always need a theoretical underpinning. For instance, creating a new medical simulation device or developing a software tool for medical education may focus more on technical specifications, user requirements, and development processes rather than theoretical considerations.

However, in designing new teaching tools or techniques, the process of creating that thing is likely enhanced by considering its uses through the lenses of various theories of learning. See also our ideas on theory-based AI.

In some cases, research may be conducted to evaluate a policy or program's effectiveness without necessarily relying on an underlying theory. Such studies might focus on outcomes and impacts rather than the theoretical mechanisms driving those outcomes. For example, an analysis of the effect of duty hour restrictions on technical competency in surgical training programs could look only at numerical data to examine the outcomes. But, again, a theory could be useful in the analysis phase and may lay the groundwork for testing potential interventions in future studies.

These are high-level abstract theories, like constructivism, realism, positivism, critical theory, etc. Many of them could be classified as paradigms, about which we have previously written. Researchers can spend an entire career wrestling with just one of these. They are so big and overarching, that they are often quite difficult, if not impossible, to apply to a scholarly project in such a practical field.

However, these grand theories often house the macro and micro theories that are easier to work with.

And, as one applies these conceptually-smaller frameworks, they bring with them the assumptions and principles built into the grand theory under which they each nest.

For example, the macro/micro theory "intersectionality" explores how a person’s social and political identities might combine to create unique instances of discrimination and privilege. Intersectionality is an offshoot of Critical Theory, which seeks to understand, critique, and challenge societal structures that perpetuate inequality and oppression. So, if choosing to apply intersectionality, one is inherently also adopting the perspectives of critical theory that look to question the status quo and facilitate emancipation.

These are frameworks that examine social processes and phenomena, like cultural norms, large-scale institutions, and broad behavior patterns. They often involve the study of social systems, political and economic structures, and historical and cultural contexts.

To continue our example of intersectionality as a theory, it can be used as both a macro and micro theory. As a macro theory, intersectionality can be used examine to broad patterns of power and inequality that affect groups of people and inform policy-making and institutional practices. For instance, looking at demographics of medical school leadership around the U.S. and discussing how the representation, or lack thereof, of various groups in positions of power, may be influencing learning outcomes for students who do not see their identities reflected in their leaders.

An example of an almost exclusively macro theory, one could look at systems theory, which views medical education as a piece of a larger healthcare system, emphasizing the interconnectivity and interdependence of different components. This could be employed to examine how healthcare reforms effect medical training, or how training in a government system (e.g. VA hospital) compares to training in a for-profit private hospital.

These frameworks examine phenomena that act at the individual or small group levels. They are concerned with the cognition, behaviors, perceptions, emotions, and actions of people alone, or in interaction with a small number of others.

Studies using micro theories often try to dissect the thinking, feelings, behaviors, emotions, communication, and relationships of subjects. They are useful for examining the nuances of individual experiences, cognitive processes, interpersonal relationships, and small-group dynamics.

Continuing our cross-cutting example of intersectionality, this theory could be applied at the individual level to examine how overlapping identities (such as race, gender, class, sexual orientation, disability, etc.) affect individual experiences, interactions, and personal narratives.

For instance, one could examine how different students with unique identities experience caring for patients at a safety-net public hospital vs. an affluent private hospital, and how their personal backgrounds inform those experiences.

Some other theories that are exclusively micro theories include cognitive load theory which examines the amount of mental effort required to perform specific tasks, and motivation theories (of which there are many) which deal with the various factors that influence adults’ choices to engage in an activity or task.

Remember, the phenomena we are studying in this field are incredibly complex. They may not fit neatly into a box, so you might need to build a new box to fit it all in, and that might mean one theory to explain part, and another theory to explain another, and your work is to tie it all together.

Just as your problem may not fit one theory, each theory may not perfectly map onto your problem. When you read about the theories you are considering, and part of a theory is resonating with you, and helping you to see your problem in a new light, follow that feeling.

If there are extraneous pieces to the theory that are making you worry it might not be right, you don’t need to include them in how you apply the theory. However, in your manuscript, you will need to tell your readers what you are excluding and why, so they can follow along with your logic.

The latest and greatest from MedEdMentor gets right at the heart of this task and is the most direct way of finding theories that can be useful to you in building your study and analyzing the results. We have a short Lesson about this tool and how it works. The bottom line is that it uses artificial intelligence to suggest real theories, derived from actual papers in the medical education literature, that may have relevance to your work. All you have to do is write a short explanation of the problem that you are trying to address and the Theory Suggester will do the rest.

Another unique resource from MedEdMentor is our medical education-specific literature search. This also uses natural language processing, so you don’t have to know any of the amazing skills librarians have to generate good search results. Just type in the problem you care about, press return, and you will receive a list of papers relevant to your work.

While this method has the benefit of providing you with references and manuscripts that you can read, you will have to sort through them to find those with theoretical framing that resonates with you.

The first medical education-specific GPT (available to ChatGPT Plus users) is able to suggest theories, just like the Theory Suggester. Its advantages over the Theory Suggester are, like other versions of ChatGPT, it can provide context for its suggestions and engage in a dialogue about the output that you receive.

However, again, like other versions of ChatGPT, we don’t exactly know where the answers are coming from. My experiences with MedEdMentor AI have always been great, but hallucination is a known issue with these platforms, so any results from this process will need to be scrutinized and verified. Everything coming from the Theory Suggester are real theories from real papers.

After you have suggestions for theories that might work for your phenomenon of interest, check our Theory Database to see if we have summaries created for those that you would like to learn more about. There, you will get a short synopsis of the framework and how it applies to medical education. You will also find resources that describe the theory in more detail, as well as papers that demonstrate the application of the theory to a problem in the field.

Lastly, each theory summary page has a section that provides a list of “similar and related theories.” If the theory that you have just read about does not seem to suit your purposes, perhaps some of the frameworks that are conceptually alike might be better fits.