Should a Dissertation Be Designed as One Study or Several Connected Studies?

Does a Dissertation Need Several Studies to Be Doctoral?

Doctoral researchers sometimes begin planning with a structural assumption: a dissertation should contain three studies, perhaps one qualitative, one quantitative, and one intervention. The number varies, but the underlying idea is similar. More studies appear to signal more substantial research.

Other dissertations develop one problem through a single extended ethnography, experiment, archival investigation, theoretical analysis, engineering project, longitudinal study, computational investigation, or case study.

Neither architecture is inherently superior. The useful question is what evidence and sequence of inquiry the dissertation's central contribution requires. The studies should follow from that logic rather than from an arbitrary target number.

Let the Research Question Determine the Number of Studies

Design the Dissertation Around an Argument, Not a Study Count

One study can contain substantial doctoral work

The word “study” can be misleading because studies differ enormously in scale and intellectual demand. A multi-year ethnography is one study. So is a complex randomized experiment, a major archival investigation, development and validation of a computational system, or a longitudinal analysis of a large dataset.

Counting studies therefore tells you remarkably little about the amount or quality of research.

A one-study dissertation may be appropriate when a single coherent design can investigate the central question deeply enough to support the original contribution expected by the doctorate.

Several studies make sense when the argument has several necessary evidentiary steps

Multi-study dissertations are particularly useful when one investigation logically informs another.

A first study might characterize a phenomenon. A second might develop or refine a measure based on what was learned. A third might test a theoretically motivated relationship using that measure. Alternatively, one study might generate a hypothesis and another test it in a different context.

Here the studies are not merely neighbors. Each performs a distinct function within a cumulative argument.

The central question should provide coherence

A dissertation with several studies needs more than a common topic. Three investigations involving generative AI, for example, are not automatically one coherent doctoral project merely because all three mention generative AI.

Ask what cumulative claim becomes possible when the studies are considered together. How does Study 2 respond to something Study 1 cannot establish? Why does Study 3 exist? What would be missing from the doctoral contribution if one study were removed?

If each study could be replaced by an unrelated project on the same broad topic without changing the dissertation's logic, the architecture may lack integration.

Do not invent studies merely to create symmetry

There is something aesthetically satisfying about “Study 1, Study 2, Study 3.” Unfortunately, aesthetic satisfaction is not a research method.

A common form of unnecessary expansion is adding a qualitative study because the main project is quantitative, or adding a survey before an experiment because dissertations are assumed to require phases. Mixed or multi-method research should be justified by the evidence the question requires.

If the extra study does not change what the dissertation can credibly conclude, its existence deserves scrutiny.

Several studies multiply project management demands

Each study can require its own literature, protocol, ethics considerations, recruitment, data management, analysis, interpretation, and writing. Different studies may also require different methodological expertise.

This means a three-study dissertation is not necessarily one project with three datasets. Operationally, it can behave like several research projects sharing a submission deadline.

This matters when deciding how ambitious the dissertation question should be. Additional studies should earn their place by strengthening the central contribution enough to justify their workload.

Sequential studies create dependency chains

Suppose Study 2 can begin only after Study 1 has identified a set of constructs, and Study 3 uses an instrument developed in Study 2. This creates a coherent progression, but also a dependency chain.

If Study 1 recruitment takes six months longer than expected, Study 2 starts late. If Study 2 fails to produce a usable instrument, Study 3 may need redesign. The intellectual integration that makes the sequence attractive also creates completion risk.

A sequential dissertation should therefore distinguish productive dependence from unnecessary dependence. Some sequencing is intrinsic to the contribution. Other dependencies can sometimes be removed by preparing later-stage infrastructure early or designing later studies so they remain meaningful under several plausible outcomes.

Connected does not have to mean sequential

Several studies can address complementary parts of a problem without one depending on the results of another.

For example, one study might examine behavior using system data while another investigates participant experiences through interviews. If both address the same theoretical problem, they may collectively support a richer argument without requiring Study 1 to finish before Study 2 begins.

This architecture can reduce cascading delays, although it creates a different challenge: integration. The dissertation must explain how the evidence fits together rather than simply presenting two separate sets of findings.

One study can also become dangerously overloaded

Avoiding multiple studies does not justify forcing every question into one enormous design.

Suppose one study contains an intervention, repeated measurements, interviews, system logs, observations, several participant groups, and institutional comparisons. Calling it “one study” does not make it simple.

Study count is therefore a poor measure of scope in both directions. Evaluate the actual research obligations.

Dissertation format matters

Some institutions permit or encourage article-based, manuscript-style, or thesis-by-publication dissertations. Others use monograph formats or impose specific rules concerning published work, co-authorship, candidate contribution, introductory synthesis, and concluding integration.

The European University Association notes substantial diversity across European doctoral programs and structures, and institutional rules governing thesis formats vary correspondingly.

Do not infer from seeing three-paper dissertations elsewhere that your program requires or even permits the same architecture. Verify the current regulations of the degree-granting institution.

Publication plans can influence architecture, but should not dictate it

A multi-study dissertation can produce several manuscripts naturally when each study makes a distinct contribution. That may be attractive for doctoral researchers building a publication record.

But adding studies solely to increase paper count can fragment the dissertation. Publication strategy should remain subordinate to the research logic and doctoral requirements.

The central test is whether each study exists because the dissertation needs its evidence, not merely because another manuscript would be useful.

Null or unexpected findings should not destroy the sequence

Sequential designs can become fragile when later studies assume that earlier studies will produce a specific result.

Imagine Study 1 is designed to establish an association and Study 2 will investigate the mechanism responsible for that association. What happens if Study 1 finds little credible evidence that the association exists?

The dissertation may still be scientifically informative, but Study 2's original rationale could disappear.

A stronger architecture anticipates plausible outcomes. Later studies may be designed around resolving uncertainty rather than requiring a preferred result to materialize.

Several studies should produce more than the sum of their papers

The dissertation-level contribution should explain what is learned by considering the studies together. This synthesis may reveal a mechanism, develop a theoretical account, demonstrate boundary conditions, integrate forms of evidence, or establish a sequence of findings that no individual study could support alone.

Without that integration, the dissertation risks becoming a folder containing several competent papers rather than one coherent doctoral argument.

Deciding Whether One Question Needs Three Studies

If later theory suggests that disciplinary context should fundamentally change how AI-supported revision operates, a third comparative or replication study may become justified. The important difference is that the study now performs identifiable intellectual work.

More Studies Do Not Automatically Mean More Doctoral Contribution

Give Every Study a Job in the Dissertation

For each proposed study, complete one sentence: “The dissertation needs this study because without it, we could not establish ______.”

If the blank is difficult to fill, the study may be optional.

Then map the dependencies among studies. Which can begin immediately? Which require findings, instruments, samples, or materials produced earlier? What happens if one study is delayed or produces an unexpected result?

This makes the connection between dissertation architecture and completion risk visible before the project becomes committed to a fragile sequence.

Decide Whether Your Dissertation Needs One Study or Several

Frequently Asked Questions About Multi-Study Dissertations

Use as Many Studies as the Contribution Requires, and No More

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.