03 · What You Need to Know
Count the Demands of the Questions, Not Just the Questions
A research question can be deceptively small on paper. One sentence may require an additional construct, another instrument, a different subgroup, a larger sample, several statistical models, another theoretical argument, or substantial qualitative analysis.
For that reason, the number of questions alone is a poor measure of study complexity.
Two demanding secondary questions could create more methodological burden than six tightly related descriptive questions. Conversely, a large multi-component project may legitimately investigate several outcomes or mechanisms because those questions were built into its design from the beginning.
The Primary Question Should Remain Identifiable
Where a study uses a primary-secondary hierarchy, that hierarchy should mean something. The primary question typically expresses the study's central objective and may determine major design decisions such as the population, sample-size calculation, primary outcome, measurement schedule, and principal analysis.
Secondary questions then provide additional information. They may examine other outcomes, mechanisms, moderators, contextual factors, or related aspects of the main problem.
In randomized trials, for example, CONSORT 2025 distinguishes the prespecified primary outcome considered most important from secondary outcomes used to assess additional effects. It also recognizes that trials often contain several outcomes while cautioning that multiple primary outcomes can introduce interpretive and multiplicity problems.
Not every methodology organizes research this way. Qualitative, mixed-methods, exploratory, and interdisciplinary studies may have several complementary questions without declaring one statistically primary. The underlying diagnostic still applies: can a reader identify what problem the study is fundamentally trying to understand?
Secondary Questions Should Form a Family, Not a Collection
A coherent set of secondary questions usually has an intelligible relationship to the central problem. One might explain a mechanism, another examine variation across contexts, and another investigate a consequence relevant to interpreting the main finding.
Contrast that with a collection of questions joined mainly because they concern the same broad topic. A study of generative AI in higher education could easily accumulate questions about student adoption, academic integrity, faculty workload, institutional policy, assessment redesign, AI literacy, employability, and environmental costs. All concern generative AI. They do not automatically constitute one investigation.
Coherent breadth
Several questions examine connected dimensions of one defined research problem and contribute to an integrated interpretation.
Scope accumulation
Additional questions are attached because they are interesting, convenient, or topically related, even though their scientific purposes increasingly diverge.
Every Question Creates an Evidence Obligation
Once a question appears in the study, researchers implicitly promise to answer it with appropriate evidence. That obligation should constrain how many questions are included.
A secondary question may require another variable, instrument, participant group, observation period, interview prompt, data source, or level of analysis. If the required evidence is not available or is collected inadequately, the question has not become more defensible merely because it was labeled secondary.
This is particularly important when additional questions are added late in planning. A measure chosen for the primary question may be poorly suited to a different construct. A sample adequate for the primary analysis may provide very imprecise estimates for subgroup or interaction analyses.
Sample Adequacy Does Not Automatically Transfer to Every Question
In quantitative research, sample-size planning is often tied to the primary question or outcome. Secondary analyses may involve different effect sizes, outcomes, subgroup sizes, interaction terms, missingness, or model complexity.
A study that is adequately powered for its primary comparison is therefore not automatically adequately powered for every secondary question. Some secondary analyses may appropriately be treated as exploratory or interpreted with greater uncertainty. Others may be sufficiently important that their sample-size requirements need to influence the design.
Multiple primary outcomes create additional statistical complications. Depending on how study success is defined, researchers may need to address multiplicity, power, and sample size differently. The precise statistical solution depends on the design rather than on a simple rule about how many questions are allowed.
Qualitative studies face an analogous constraint. Adding questions may broaden the phenomena, contexts, or participant experiences that must be represented. Eventually, the available interviews, observations, documents, or cases may no longer provide sufficient depth for each analytical claim.
Secondary Questions Increase Analytical Burden
More questions usually mean more analytical decisions. Quantitative studies may accumulate outcomes, predictors, subgroups, interactions, time points, models, and comparisons. In inferential settings, this may introduce multiplicity concerns and increase the opportunity for chance findings if many tests are conducted without an appropriate analytical strategy.
Recent methodological reviews of multiple outcomes in randomized trials continue to emphasize prespecification, outcome classification, study power, sample size, and the relationship between research questions and outcome selection. These issues are particularly developed in clinical research, but the broader lesson is useful elsewhere: additional questions have consequences for design and interpretation.
In qualitative research, analytical burden takes another form. More questions can generate additional coding structures, comparisons, interpretive frameworks, and bodies of literature. The problem may become insufficient analytical depth rather than excessive statistical testing.
Participant Burden Can Set the Limit Before Analysis Does
Researchers sometimes treat an additional question as inexpensive because the analysis itself seems straightforward. The participant may experience the cost differently.
Another question might require another questionnaire scale, interview topic, laboratory procedure, follow-up visit, diary entry, or assessment. Individually these additions can appear minor. Collectively they can lengthen instruments, increase fatigue, reduce completion quality, contribute to attrition, or create unnecessary burden.
The scientific value of secondary questions should therefore be considered against what participants are being asked to contribute.
The Literature Review Can Reveal That the Study Is Splitting Apart
A practical diagnostic is to outline the literature required for each question. If every secondary question can be justified using the same central conceptual foundation with modest extensions, the study may remain coherent.
If each question needs several pages of largely unrelated literature, a different theoretical framework, and its own research gap, the project may be functioning as several studies.
The same diagnostic can be applied to the discussion. Would the findings be interpreted together, or would the manuscript need a separate discussion for each question with little intellectual interaction among them?
Watch for the Question That Starts Competing With the Primary Question
The problem sometimes becomes visible before the total number of questions seems excessive. One particular secondary question begins consuming disproportionate attention.
It may require a larger sample, an additional population, substantial new measures, or a separate analytical framework. At that point, the important issue is not the number of secondary questions but whether a secondary question is distracting from the primary question.
A study with six modest secondary questions may remain focused, while a study with only one secondary question may become distorted if that question effectively introduces another primary investigation.
Do Not Use Available Data as a Reason to Keep Expanding the List
Large datasets and comprehensive surveys make additional questions especially tempting. Once a variable is being collected anyway, another analysis may seem almost free.
It is not. Even without additional data collection, a question creates obligations of rationale, analysis, interpretation, and reporting. Researchers should therefore resist adding questions simply because the necessary data will be available.
Exploratory Questions Still Count Toward Study Complexity
Calling a question exploratory does not make its demands disappear. Exploratory questions may legitimately be more open-ended and hypothesis-generating, but they still require evidence and analytical attention.
Researchers can specify worthwhile exploratory questions prospectively, particularly when doing so improves measurement or sampling. The same scope test applies: does the study have enough capacity to investigate them responsibly?
There Is No Defensible Universal Number
Advice such as "a study should have no more than three research questions" may be convenient for teaching, proposal templates, or particular institutional contexts, but it should not be mistaken for a universal methodological rule.
Research designs differ too much for such a number to travel well. A small master's project, a qualitative case study, a multicenter trial, a national cohort, and a mixed-methods evaluation operate under very different constraints.
Watch Out
If a supervisor, institution, funder, ethics committee, journal, or degree program imposes a practical limit on research questions, follow the applicable requirement. That local rule should be distinguished from a general methodological claim that the same number is appropriate for every study.
The Boundary May Be a Research Program Rather Than a Bigger Study
If several questions are individually important but cannot be investigated adequately within one project, the solution is not necessarily to discard them. They may form a sequence of connected studies.
One project could establish the phenomenon, another investigate mechanisms, and a later study test an intervention. As the questions become linked across multiple investigations, the work may begin to resemble a research program rather than a single study.
Scientific ambition does not require methodological overcrowding.
