%202.png)
Research Design and Methodology Chapter
Turns a methodology chapter that describes what you did into one that justifies every choice against the alternative it beat.
Free. Works on Claude, ChatGPT, Gemini or any assistant that accepts a skill file.
What this skill does
The method, encoded.
"Describes without justifying" is the most common criticism written on methodology chapters, and it is not a stylistic complaint. A chapter that says the study used semi-structured interviews with twelve participants recruited through the researcher's network has said what happened and nothing about whether it should have. Every question an examiner will ask is one the chapter did not answer.
This skill rewrites the chapter as a sustained justification. It starts by establishing what your department actually expects, because methodology conventions are local and a chapter written to a generic template gets marked against a specific one. It gives honest guidance on how much philosophical positioning your discipline requires, with a test that removes the padding: if the design would be identical under the opposite position, the sentence is decoration.
It then separates target population, accessible population, frame and achieved sample so the gaps become visible, names a convenience sample as one and bounds the claims accordingly, requires the adoption of an existing instrument to be justified as thoroughly as building one, specifies the analysis rather than labelling it, and insists on a single coherent quality vocabulary rather than mixing validity and trustworthiness language. Limitations sit next to the choices that create them.
Best used for
- Turning a described methodology into a justified one
- Deciding how much epistemology a discipline requires
- Naming a convenience sample honestly and bounding the claims
- Justifying adoption of an existing instrument
- Choosing between validity and trustworthiness vocabularies
- Revising the chapter post-fieldwork to report what happened
Typical inputs
What you give it.
Final research questions or hypotheses, The design decisions actually made and the procedure actually followed, Departmental requirements and the marking rubric for the chapter, The conceptual framework and operationalisation table, Instrument source and validation evidence where adopted, A fieldwork log recording deviations from plan
Typical outputs
What you get back.
A subsection plan with a justification question per heading, A proportionate philosophical positioning section with stated design consequences, Layered approach, strategy and time horizon sections with rejected alternatives, A four-level population and sampling table with gaps and claim consequences, An instrument justification with a construct-to-item mapping, A reproducible procedure section including deviations, An analysis specification mapped to research questions, planned versus exploratory, A single coherent quality-criteria section with practices attached, Design limitations stated with cause and consequence, A completed coherence audit across the justification chain
Method coverage
What the skill works through.
- Why methodology chapters lose marks
- Finding out what your department actually requires
- Turning every heading into a justification question
- How much research philosophy is enough
- Approach, strategy, time horizon and the alternatives you rejected
- Target population, accessible population, frame and sample
- Naming a convenience sample without wrecking your study
- Justifying an adopted instrument as carefully as a built one
- Writing a procedure a stranger could repeat
- Specifying the analysis instead of naming it
- Validity and reliability, or credibility and transferability, but not both
- Putting limitations where the choice was made
- The coherence audit from question to limitation
Download
Free skill. One file.
Enter your email once. Every skill you download after that takes a single click.
How to install
Add the skill file and the five kernel protocols to a Claude Project, a ChatGPT Project, a Gemini Gem, or paste them at the top of any assistant conversation. Then give it your real research material, not a description of it.
Download skillQuestions
Common questions.
How much research philosophy should a masters methodology chapter contain?
It depends genuinely on the discipline, and both extremes are defensible within their own conventions. The reliable test is consequence: every philosophical claim should connect to at least two design decisions, and if the design would be identical under the opposite position, the passage is decoration. Check two recent accepted dissertations from your own department, which tell you more than any general guidance.
How do I justify a convenience sample?
Name it as a convenience sample, state why it was the available option given access and timeline, and then state precisely what it means for the claims: the findings describe this sample, the design does not support generalisation to the population, and any wider claim rests on an argument about similarity rather than on statistics. Also state the direction of the likely bias. Naming it costs you a narrower claim; concealing it costs credibility across the whole thesis.
Do I need to justify using an existing questionnaire?
Yes, and it is the justification most often skipped. You need the source, evidence of validity and reliability in prior use, the population it was developed on and whether that matches yours, whether it measures your constructs as your framework defines them, permission where required, and why it beat the alternatives. An instrument validated on a different population may need re-validation, and saying so is a strength.
Should I use validity and reliability or credibility and transferability?
One or the other, chosen to match the design, and never both. They come from different traditions and rest on different assumptions about what a finding is. A chapter that claims external validity and transferability together signals that neither concept was understood. In mixed designs, apply each vocabulary to its own strand and say so explicitly.
Where should limitations go, the methodology chapter or the conclusion?
Both, doing different jobs. Design limitations belong in the methodology, next to the choice that creates them, each stated with its cause and its specific consequence for interpretation. The final chapter then revisits limitations in the light of what was actually found, which is a shorter and different discussion. Generic lists like "small sample, more research needed" gain nothing in either place.
What if my analysis plan changed after I saw the data?
Report the change, and distinguish planned analyses from exploratory ones. Exploratory analysis is legitimate and common; presenting it as pre-planned is not. Writing the analysis section in full before data collection and lodging it with your supervisor makes this distinction defensible rather than merely assertable.
The skill chain
Works well with.
Research where people already are.
Analyse it where you already work.
Yazi helps researchers conduct surveys, AI interviews and longitudinal research directly through WhatsApp.
