A dissertation research design explains how the study will produce evidence capable of answering the research question. It connects the study purpose to the overall approach, evidence source, selection process, data collection and analysis.
Start with the research question
The design should follow from what the study needs to find out. If the question is still unstable, begin with research question help.
Different questions imply different designs. A question asking how participants experience a process may require an interpretive qualitative design. A question asking whether two measurable variables are associated may require a quantitative correlational design. A question asking both how a pattern occurs and how participants explain it may justify a mixed-methods design.
This is why design should be chosen after the question is sufficiently clear. Starting with “I want to use interviews” or “I want to use SPSS” reverses the logic: it allows a preferred tool to determine the problem instead of allowing the problem to determine the evidence.
Distinguish design from methods
Research design is the overall logic of the study. Interviews, questionnaires, datasets, documents or observations are methods or evidence sources within that design.
Choose an approach that fits the evidence
Qualitative, quantitative and mixed-methods approaches serve different purposes. The choice should follow the question, evidence need and claims the study expects to support.
The design should state what type of evidence is needed and why that evidence can answer the question. Qualitative designs are often appropriate when context, meaning, process or experience is central. Quantitative designs are useful when the study needs measurement, comparison, association or estimation. Mixed-methods designs need an additional justification for why neither evidence type is sufficient alone and how the two strands will contribute to one answer.
The label is only the beginning. Two qualitative dissertations can differ substantially in case selection, data collection and interpretation, just as two quantitative dissertations can use completely different sampling strategies and statistical procedures.
Connect selection, collection and analysis
The selected evidence should be collectable and analysable in a way that answers the question. Planning these decisions separately often creates a methodology that is detailed but not coherent. The planned evidence should also be compatible with the dissertation data analysis needed to answer the question.
A coherent design can be traced as a chain. The research question determines the required evidence; the sampling strategy determines where that evidence will come from; the collection method determines how it will be obtained; and the analysis determines how the evidence will be converted into findings. A weakness at one stage affects the stages that follow.
For example, a dissertation cannot plan a statistical comparison between groups if the sampling and questionnaire do not produce usable group-level measurements. Likewise, a qualitative study cannot promise detailed interpretation of participant experience if its collection method produces only short closed responses.
Account for feasibility, ethics and limitations
A design may be suitable in principle but impractical if participants, records, permissions, time or data are unavailable. Ethical requirements and limitations should be tied to the actual design. Where the study involves participants, identifiable information, sensitive evidence, or approval requirements, review the dedicated dissertation ethics guidance.
Feasibility is part of design quality rather than an administrative issue added at the end. A project that requires access to records the student cannot obtain, a participant group that cannot realistically be recruited, or repeated observations that cannot be completed within the available period is not a workable design even if it looks theoretically attractive.
The final design should therefore balance methodological fit with access, time, skills, permissions, ethics and available evidence. Limitations should describe the consequences of those decisions rather than simply stating that the project had “limited time and resources.”
For whole-chapter integration, return to Methodology Chapter Help.