The methodology chapter is where your dissertation proves it is rigorous scholarship, not just an opinion or anecdote. Your committee scrutinizes it more closely than any other chapter because methodology determines whether your findings mean anything. If your design is flawed, your data collection is biased, or your analysis is invalid, the entire dissertation falls apart. This guide covers the most common methodology mistakes doctoral students make, what your committee is actually looking for, and how to write a methodology section that passes defense.
What makes a strong methodology chapter
A strong methodology chapter does five things:
- Justifies the design choice: Explains why this design (quantitative, qualitative, mixed methods) is appropriate for the research question.
- Describes the design with precision: Specifies exactly what you're doing (quasi-experimental pre-post, phenomenology, grounded theory, survey) and why it fits the question.
- Details the sample/participants: Who are they, how many, how were they selected, what are inclusion/exclusion criteria?
- Explains data collection: What instruments or methods, how long, when, where, who administers it?
- Describes analysis: What statistical tests (ANOVA, regression) or qualitative analysis method (thematic, narrative) will you use and why?
Committees often reject methodology sections because students skip the why and only provide the what. They describe the design but don't justify it. They list methods but don't defend them. A strong methodology answers both questions for everything you propose.
Common methodology mistakes
1. Design doesn't match the research question
Question: "What is the lived experience of nurses returning to school?" Design: "I will administer a 50-item Likert scale survey." This is a mismatch. Lived experience requires qualitative methods. A Likert scale measures agreement with statements, not experience. The committee will immediately flag this. Ensure your design actually addresses your question. Experiential questions need qualitative approaches. Effectiveness questions need experimental or quasi-experimental designs. Prevalence questions need surveys.
2. Sample is too small or selection method is biased
You propose n=20 for a quantitative study. Too small to detect effects or generalize. Or you recruit "volunteers from a local clinic" without explaining how you'll avoid selection bias (are these the healthiest, most motivated patients?). Sample size must be justified. For quantitative studies, cite power analysis. For qualitative, justify why 10–15 in-depth interviews is sufficient for saturation. Address selection bias explicitly.
3. Data collection methods are vague or unvalidated
You propose to collect data via "interviews and observation." How long are interviews? How many? Will you use an interview guide? Is observation structured or unstructured? For observation, describe a protocol. If using an existing instrument, cite it and provide psychometric evidence (reliability, validity). If creating your own, explain how you'll pilot test it. Vague methods make your study unreplicable.
4. Analysis plan is missing or unclear
You propose a quantitative study but say only "data will be analyzed using SPSS." Committees need detail: Will you use descriptive statistics? Which inferential tests? Why those tests? How will you check for normality, homogeneity of variance, etc.? For qualitative, "data will be analyzed for themes" is not enough. Describe your coding process: How many coders? How will you ensure reliability? Will you use software (NVivo, Atlas.ti)? What is your coding scheme? Absent analysis plans are major red flags.
5. IRB requirements not addressed
You propose to collect data from a vulnerable population (minors, incarcerated people, patients with severe mental illness) but don't explain how you'll get informed consent, maintain confidentiality, or manage risk. Or you propose to access private health information but haven't addressed HIPAA compliance. Committees expect you to identify IRB concerns proactively and explain how you'll address them. If your study needs IRB approval, say so. If it's exempt, explain why (e.g., "De-identified secondary data analysis; exempt under 45 CFR 46.104(d)(4)(ii)").
Methodology section by design type
| Design | Key elements to include | Common weakness |
|---|---|---|
| Quantitative (RCT, quasi-experiment, survey) | Design type, sample size justification, randomization/assignment method, control variables, instruments with validity/reliability evidence, statistical tests with assumptions checked | Vague description of which group gets what, missing power analysis |
| Qualitative (phenomenology, grounded theory, ethnography) | Design choice and justification, participant recruitment/selection, sample size and saturation criteria, data collection method, coding procedure, trustworthiness strategies (triangulation, member checking) | No explanation of how saturation will be achieved; missing credibility/reliability checks |
| Mixed methods | Overall design (convergent, explanatory, exploratory), qual + quant components described separately with full detail, how qual and quant phases connect, integration/mixing strategy | Treating it as "qual and quant" without explaining how they're integrated or which is primary |
| Quality improvement / program evaluation | Theory of change or logic model, outcome measures (process + outcome), baseline data collection, implementation timeline, fidelity monitoring, analysis of before/after data | No logic model; missing process measures; no fidelity check; only measuring outcome without process |
Working with a methodology consultant
If your committee gave methodology feedback and you're unsure how to address it, a methodology consultant can help you:
- Translate committee feedback into concrete design changes
- Justify design choices in writing
- Address validity and reliability concerns
- Develop a realistic analysis plan
- Prepare for IRB submission
Methodology consulting usually costs $75–150/hour and takes 3–5 hours to work through a methodology section. It's worth the investment if your design is weak or committee feedback is complex.
Your methodology committee is questioning your design?
Tell us what feedback you received. We'll help you strengthen the section and address concerns before you resubmit.
Get methodology help Back to proposal helpRelated guides
Frequently asked questions
Not necessarily. Most dissertations place lengthy instruments (full survey, interview guide, observation protocol) in the appendix and reference them in the methodology section. Your chapter should describe the instrument and cite it, but readers can see the full text in the appendix. This keeps your methodology section readable while providing full transparency.
That's fine. Exploratory designs don't require hypotheses — they ask research questions instead. Your methodology should frame it as exploratory, justify why (e.g., "prior research is limited; an exploratory approach allows investigation without constraining findings to pre-specified hypotheses"). Just be clear that you're exploring, not testing.
Very detailed. You should specify: Which variables go where? What's your dependent variable? Your independent variable? How will you handle missing data? What's your alpha (significance level)? Will you do univariate or multivariate analysis? What are your assumptions checks? The more specific, the more defensible. You're telling your committee exactly what you'll do before you see the data — that's research rigor.