How Do You Prevent an Interdisciplinary Question From Becoming Impossibly Broad?

How Do You Keep an Interdisciplinary Question From Trying to Explain Everything?

You begin with a focused problem. Then another discipline reveals an important mechanism. A third points to a contextual factor you had overlooked. Someone raises policy. Someone else raises ethics. Before long, your research question seems to require five theories, several levels of analysis, multiple populations, and a methodological plan that would keep a small research institute occupied for years.

The additions may all be intellectually defensible. That is precisely what makes interdisciplinary scope difficult.

Real-world problems rarely stop at disciplinary boundaries, but an individual study must stop somewhere. The challenge is to narrow the question without removing the relationships across disciplines that made an interdisciplinary approach necessary in the first place.

Narrow the Explanatory Task, Not Necessarily the Number of Disciplines

Interdisciplinary Scope Expands Unless You Give It Boundaries

Complex problems are larger than individual research questions

Interdisciplinary research is often motivated by problems that exceed the scope of a single discipline. The National Academies defines interdisciplinary research around the integration of knowledge from multiple disciplines to advance understanding or address problems whose solutions extend beyond a single disciplinary area.

That definition does not imply that one study must explain the entire complex problem.

Consider educational inequality. A comprehensive account might involve learning processes, family resources, institutional structures, economics, public policy, geography, technology, culture, health, and historical conditions. Each can matter. A doctoral thesis, journal article, or ordinary funded project cannot investigate all of them with equal depth.

The larger problem provides the context. The research question identifies the particular piece of that problem your study can reasonably investigate.

Do not mistake relevance for inclusion

Interdisciplinary researchers encounter a recurring problem: once you recognize that several systems interact, almost everything begins to look relevant.

Suppose you are studying students' reliance on generative AI. Learning theory may be relevant. So might psychology, human-computer interaction, information systems, sociology, ethics, policy, assessment, digital inequality, privacy, and labor economics.

But a factor being relevant to generative AI in education does not mean it belongs in your research question.

This distinction is one of the most useful defenses against uncontrolled scope. Ask whether each proposed disciplinary contribution is necessary to the answer, not merely whether somebody could write a convincing paragraph explaining why it matters.

Give the question one primary job

Broad interdisciplinary questions often contain several research projects disguised as one sentence.

For example:

How do technological, psychological, pedagogical, social, ethical, and institutional factors influence university students' adoption, use, learning outcomes, academic integrity, and attitudes toward generative AI?

The difficulty is not simply that the sentence is long. It contains multiple outcomes, explanatory levels, mechanisms, and disciplinary perspectives. Answering one part does little to resolve the others.

A more focused question might ask:

How do characteristics of AI-generated explanations and students' metacognitive judgments interact to influence whether students accept incorrect academic feedback?

The revised question remains interdisciplinary if those technological and learning-related mechanisms genuinely require different bodies of knowledge. But it now has one principal explanatory job.

Define where the disciplines actually meet

A useful interdisciplinary question does not merely contain several disciplinary ingredients. It identifies an interface.

Perhaps the interface is between system behavior and human judgment, biological processes and social conditions, individual behavior and institutional rules, environmental change and economic decision-making, or technology design and classroom practice.

Once you identify that interface, scope the study around it.

This follows a broader principle in interdisciplinary research: integration matters more than disciplinary accumulation. The National Academies has specifically cautioned against research plans that simply staple together overlapping disciplinary components rather than integrating knowledge and skills around a coherent problem.

If you cannot identify where the disciplines need one another, revisit whether the question actually requires multiple disciplines.

Limit the number of explanatory layers

Scope often becomes unmanageable when a question tries to connect too many levels simultaneously.

You might begin with individual learner behavior, then add teacher practices, institutional policy, technological design, national regulation, and socioeconomic inequality. Each level can plausibly affect the others, but examining the entire chain may require different data, units of analysis, theories, sampling strategies, and methods.

Choose which cross-level relationship is central to the study.

For example, a study might investigate how institutional assessment rules influence individual students' use of generative AI. That already connects institutional and individual levels. Adding national AI policy may be unnecessary unless the research question specifically concerns how policy changes institutional rules.

Context can be acknowledged without becoming another explanatory level.

Distinguish what you will study from what you will hold as context

Researchers sometimes expand a question because they fear that excluding a relevant factor means pretending it does not exist.

It does not.

You can acknowledge factors in the literature review, delimit the study explicitly, control or account for selected variables where methodologically appropriate, and discuss limitations without making every relevant factor a central research construct.

A study of students' trust in AI-generated feedback can acknowledge institutional AI policies without making policy an independent variable. It can recognize socioeconomic inequalities without claiming to explain them. It can discuss ethical implications without turning ethics into a separate empirical research question.

Research boundaries are not claims that the world ends at the edge of your conceptual framework.

Use the removal test to reduce disciplinary load

For every discipline, theory, construct, mechanism, and method you plan to include, ask what happens if you remove it.

This is especially useful when an interdisciplinary project has accumulated components gradually. Academic frameworks have a curious tendency to acquire variables more easily than they lose them.

Do not solve scope problems by merely shortening the wording

A question can be linguistically concise and conceptually enormous.

“How does AI transform higher education?” contains only a few words but potentially encompasses teaching, learning, assessment, governance, labor, infrastructure, ethics, equity, economics, policy, and institutional change across countless populations and contexts.

Conversely, a longer question may be well bounded if it specifies the phenomenon, population, mechanisms, context, and relationship clearly.

Judge scope by what evidence would be required to answer the question, not by word count.

Specify the population and setting only where they matter

Population and context can narrow a question substantially, but specificity should serve the research logic.

“University students” may still be too broad if the mechanism depends on disciplinary background, educational level, assessment type, or previous AI experience. On the other hand, restricting the study to one institution simply because that is where participants are accessible does not automatically create a theoretically meaningful boundary.

Ask which contextual characteristics could plausibly alter the phenomenon or the interpretation of the findings. Those deserve explicit attention. Other restrictions may be practical sampling boundaries that should be described as such.

Limit the claims along with the data

Scope is not controlled only at the research-question stage. Researchers can conduct a modest study and then make conclusions about a much larger problem.

If your data concern one interaction between two mechanisms, your conclusions should principally concern that interaction. If your sample comes from a particular institutional context, do not casually convert the findings into claims about higher education globally.

A bounded interdisciplinary study can contribute to understanding a larger problem without claiming to resolve it.

Some questions are broad because they contain several legitimate studies

Occasionally, narrowing removes something genuinely important. The problem is not unnecessary scope but the fact that the research idea contains multiple linked questions that cannot be investigated adequately within one design.

For example, understanding an intervention may require first studying its technical behavior, then how users respond to it, and later how institutions implement it. Compressing all three into one study could weaken each component.

In that situation, the better solution may be to determine whether the interdisciplinary question should be divided into several linked studies rather than forcing everything into one research question.

Feasibility is an intellectual constraint, not merely a logistical inconvenience

Time, expertise, access, data, funding, equipment, and analytic capacity affect what can be investigated rigorously. They therefore belong in question development.

A question that theoretically requires expertise in four areas but is being conducted by a researcher competent in only one deserves reconsideration. The solution might be collaboration, training, narrowing, or redesign.

Relevant expertise should ideally enter while the question remains changeable, particularly when deciding whether to build the study before collaborators from the relevant disciplines are involved.

Narrowing a Question About Generative AI Without Removing Its Interdisciplinarity

Every component has a plausible rationale. Together, however, they create several different explanatory problems.

The resulting study can remain interdisciplinary because the interaction between technological output and learner judgment still requires integration. What disappeared was not the interdisciplinarity but the expectation that one project should represent every important issue surrounding AI in education.

Common Mistakes When Narrowing Interdisciplinary Questions

Set Boundaries Around the Interaction You Actually Want to Explain

When an interdisciplinary question becomes unwieldy, do not begin by asking which discipline to delete. Begin by asking what the study absolutely needs to explain.

A useful final test is to imagine the evidence required for a convincing answer. If answering the question would require several unrelated datasets, incompatible units of analysis, a small army of specialists, and three dissertations hiding in a trench coat, the scope probably still needs work.

Before Finalizing an Interdisciplinary Research Question, Check Its Scope

Questions About Narrowing Interdisciplinary Research

Keep the Interdisciplinary Connection and Cut the Unnecessary Scope

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.