Statistical Method Recommender · 7 min read

How to defend your methodology chapter in a viva

The questions examiners actually ask about design, sampling, instruments and analysis, and how to prepare evidence-backed answers for each one.

Expect four families of questions

Almost every methodology question falls into one of four groups: why this design, why this sample, why this instrument, and why this analysis. Prepare one short, specific answer for each, each tied to your research objective rather than to convenience.

Justify the analysis with the decision trail

Run your data description through the Statistical Method Recommender and keep the output. It gives you the objective, the measurement levels, the recommended test, the assumptions and the fallback in one place — exactly the trail an examiner asks you to reconstruct verbally.

Being able to say 'the outcome is binary and the predictors are mixed, so logistic regression; the Hosmer-Lemeshow test supported fit' ends the question faster than any amount of general reading.

Have your assumption checks printed

Bring the normality, variance, multicollinearity and independence checks as an appendix. When an examiner asks whether you tested them, the answer should be a page number, not a promise.

Own the limitations before they are raised

Name your two or three real limitations — sampling frame, self-report bias, single-city scope — and state what they do and do not affect. A candidate who identifies limitations precisely reads as competent; one who denies them reads as unaware.

Rehearse with the tools open

Walk a colleague through your topic in the Research Topic Evaluator, your sample in the Sample Size Calculator and your dataset in the Data Quality Checker. If you can narrate all three without hesitating, the chapter will hold.

Chat on WhatsApp