How Do You Know When Your Study Is Trying to Answer Too Much?

When Does an Ambitious Study Become an Overloaded One?

You start with one research question. Then another interesting question appears. Perhaps you should compare two populations. Add interviews to explain the survey results. Examine several outcomes. Include another institution. Test whether the relationship differs by demographic characteristics. While you are there, why not investigate another variable that the literature says might matter?

Each addition may be defensible on its own. The difficulty appears when you consider what the study must accomplish collectively.

A complex study is not automatically too broad. Some research problems genuinely require multiple methods, populations, outcomes, cases, or phases. The warning sign is different: the project begins to contain more substantive questions than its design, evidence, resources, or conceptual framework can answer convincingly.

The practical question is therefore not simply, "How many research questions do I have?" It is: Do these questions still form one coherent, feasible investigation, or am I quietly designing several studies under one title?

Look at What Your Questions Require Collectively

Excessive Scope Usually Reveals Itself Through the Demands It Creates

Start by identifying the central question the study exists to answer

A coherent research project usually has a recognizable center. Even when it contains several objectives or subsidiary questions, you should be able to explain what central problem holds them together.

This matters because research questions influence study design, the population being studied, the data that must be collected, and the analyses required. Methodological guidance on research-question development consequently emphasizes the value of a clearly defined primary question. Multiple questions can increase design and analytical complexity, and secondary questions should not compromise the study's ability to answer its primary question.

Suppose a study asks whether students' generative AI use is associated with writing self-efficacy. A secondary question examining whether that relationship differs by prior AI experience may fit naturally within the same conceptual inquiry.

Now imagine adding questions about faculty attitudes toward AI, institutional policy implementation, students' academic-integrity violations, the accuracy of AI-generated references, and the economic cost of institutional AI subscriptions. All concern generative AI in higher education. That shared topic, however, does not make them one research problem.

A topic can contain many studies. A study needs a more specific organizing question.

Count the methodological obligations, not just the questions

Two research questions can create more work than six. What matters is what each question obligates you to do.

For every question, ask:

• What population, cases, documents, or other evidence does this require?
• What constructs or phenomena must be defined?
• What measurements or data-generation procedures are necessary?
• What comparisons must be supported?
• What analytical approach is needed?
• What sample size, information depth, or evidential coverage does that approach require?
• What expertise, access, ethical approval, time, and resources does it add?

If several questions depend on the same participants, measures, conceptual framework, and analytical strategy, they may fit comfortably within one project. If each question introduces a new methodological infrastructure, the nominal number of questions tells you very little.

Several related questions can still belong in one study

Do not mistake coherence for simplicity. A mixed-methods study may legitimately combine quantitative and qualitative questions. A longitudinal project may examine several related outcomes. A comparative case study may intentionally investigate multiple settings. An experiment may include secondary outcomes alongside a primary outcome.

The question is whether those components are methodologically and conceptually integrated.

For example, a mixed-methods study might first estimate whether students' use of an AI-supported feedback system is associated with writing outcomes, then use interviews to investigate how students experienced that feedback. The qualitative component can be justified because it addresses a related dimension of the same central phenomenon and the design specifies how the strands relate.

Contrast that with conducting a student survey, interviewing faculty about institutional policy, analyzing AI-generated citations, and auditing university expenditure simply because all four activities concern AI. That is breadth by topical association, not necessarily a coherent mixed-methods design.

One warning sign is that each question needs its own literature review

Try outlining the conceptual background required to justify each research question.

If all the questions arise from the same body of literature and theoretical problem, the project may be coherent. If Question 1 requires literature on student motivation, Question 2 requires an independent literature on faculty technology adoption, Question 3 requires research-integrity scholarship, and Question 4 requires organizational policy analysis, you may be assembling several intellectual problems.

This is not an automatic reason to separate them. Interdisciplinary research sometimes requires several bodies of literature. The important test is whether you can articulate the theoretical or analytical relationship among them rather than relying on a broad topic label to hold everything together.

Another warning sign is that the population keeps changing

A project can become overloaded when each new question introduces another population.

Imagine beginning with undergraduate students, then adding postgraduate students "for comparison," faculty members "for another perspective," administrators "because they make policy," and employers "to assess workforce relevance."

Those groups may all matter to the larger research problem. They also occupy different roles and may require different sampling strategies, instruments, ethical considerations, conceptual frameworks, and analytical comparisons.

The issue is not that a study can never contain multiple populations. It can. The issue is whether those populations are necessary to answer one integrated question and whether the design can examine their differences adequately.

If not, narrowing the scope to what can be investigated rigorously may produce a stronger project than superficially representing every stakeholder.

Too many constructs can create a study without a clear explanatory model

Another common pattern is variable accumulation. A researcher begins with two constructs, reads more literature, discovers six additional variables associated with the outcome, and decides to include all of them.

Relevance to the topic is not sufficient justification for inclusion.

Each construct should have a defensible role in the research question, theoretical or conceptual framework, and analysis. Adding variables can increase measurement burden and analytical complexity. In quantitative research, additional outcomes and comparisons may also raise issues involving statistical power, sample-size requirements, and multiple testing.

If you cannot explain why a variable must be measured beyond "previous studies have examined it," it may not belong in the current study.

Too many outcomes can blur what success or failure actually means

A related problem occurs when a study has no clear primary outcome or objective.

Methodological guidance commonly distinguishes primary and secondary objectives or outcomes because they serve different roles in study planning and interpretation. The primary question should drive the design rather than becoming one result among many unrelated analyses.

Suppose an intervention study measures examination scores, motivation, satisfaction, engagement, self-efficacy, anxiety, retention, attendance, cognitive load, and intention to continue using the technology. The study may have legitimate reasons for several of these outcomes. But if every outcome is treated as equally central, it becomes difficult to determine what claim the study was principally designed to evaluate.

A useful question is: If only one result could be known at the end of the project, which result would most directly answer the problem that justified the study?

If you cannot answer that, the study may lack a clear center.

Every added comparison increases what the study must support

Comparisons can multiply rapidly. You may want to compare first-year and senior students, public and private institutions, disciplines, gender groups, AI users and non-users, high- and low-performing students, and different AI tools.

Each comparison may sound straightforward when written as one sentence. Collectively, however, they can require sufficient representation in each relevant subgroup, appropriate analytical procedures, and a defensible rationale for interpreting the differences.

In quantitative studies, planned multiple comparisons may affect statistical testing and sample-size considerations. In qualitative research, adding comparison groups may require enough information within each group to support the intended comparative interpretation.

Comparison should therefore be driven by the research problem, not by the fact that the dataset contains another categorical variable.

The "while we're here" question is a particularly useful warning sign

Some questions enter a project not because the study needs them, but because collecting the data seems convenient.

"While we're surveying the students, let's ask about mental health."

"Since we're interviewing faculty, let's ask about job satisfaction too."

"The dataset already contains GPA, so let's see whether it predicts everything."

Exploratory analysis can be legitimate, and secondary questions can generate valuable insights. The problem arises when opportunistic questions are retrospectively presented as though they were part of the study's original central rationale or when they consume resources needed for the primary inquiry.

Feasibility is more than finishing data collection

A study is not feasible merely because you can collect the data before the deadline.

Feasibility also concerns whether you have adequate access, participants or evidence, technical expertise, time, funding, personnel, and analytical capacity to answer the question properly. The FINER framework explicitly treats feasibility and manageable scope as characteristics of a good research question.

Consider the entire research process: developing or selecting appropriate measures, obtaining approvals, recruiting participants, generating or cleaning data, conducting analyses, interpreting findings, addressing alternative explanations, and reporting the study transparently.

If the only way to complete the project is to perform each component superficially, the problem is not merely workload. The scope may be compromising research quality.

Ask whether removing a question damages the central study

This is one of the most useful tests for excessive scope.

Take each secondary research question and imagine removing it. Does the central study become conceptually incomplete? Does the primary question become impossible to answer? Does the design lose an essential explanatory component?

If yes, the question may belong.

If the project remains intellectually coherent and answers its central question just as well, the secondary question may be interesting but nonessential. It could become exploratory analysis, future research, or a separate project.

Research inevitably involves deciding what should deliberately remain outside the current inquiry. Exclusion is not evidence that the omitted question lacks value. It may simply mean that answering it properly deserves another study.

Sometimes the project really contains several studies

A revealing test is to divide your research questions into groups according to the evidence and methods required to answer them.

If one cluster concerns students and survey data, another concerns faculty and interviews, and another concerns institutional documents and policy analysis, ask whether these strands are deliberately integrated to answer a higher-order question.

If they are, a multi-phase or mixed-methods design may be justified.

If they are not, you may have several studies connected by a common topic. That can be intellectually productive. It simply needs to be recognized for what it is. The relevant design question then becomes whether the expanding scope should be separated into multiple studies.

Broad scope and important research are not synonymous

Researchers sometimes hesitate to remove questions because the smaller project feels less impressive. This can create an unfortunate equation: more variables plus more populations plus more methods equals more contribution.

Contribution does not work that way.

A focused study can make a substantial contribution if it provides convincing evidence about an important unresolved question. Conversely, an enormous project can produce many results while resolving none of its questions particularly well.

The goal is not to maximize the number of things investigated. It is to maximize how well the study answers the question it claims to answer.

When One AI-in-Education Study Quietly Becomes Four Studies

The appropriate response is not automatically to delete everything except the first question. Perhaps the research problem genuinely concerns the relationship among student practices, faculty expectations, and institutional policy. In that case, the researcher needs a design, resources, conceptual framework, and integration strategy capable of supporting that broader inquiry.

If the original purpose is simply to understand student AI use and writing self-efficacy, however, the faculty and policy questions may be worthwhile projects that do not need to be answered now.

Common Misjudgments About How Much One Study Should Answer

Audit the Study Before Adding Another Question

When you suspect that a project is becoming overloaded, write every research question in one place and map the obligations each creates. This is often more revealing than rereading the proposal prose.

If the project needs narrowing, do not remove elements randomly. Return to the central research problem and identify what evidence is necessary to answer it. Then decide which populations, variables, contexts, comparisons, and secondary questions are essential.

The resulting boundaries should be reflected explicitly when you define the study's scope and delimitations. If new questions continue to appear after the project begins, a different problem may be emerging: preventing scope creep once the research is underway.

Is Your Research Project Trying to Do Too Much?

Frequently Asked Questions About Overly Broad Research Studies

A Study Is Too Broad When Breadth Starts Weakening the Answers

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.