· By Thesisroom
Updated
How to Defend Your Sample Size When Examiners Push Back
Panels almost always press on sample size. Here is how to handle defending sample size in viva, for both qualitative and quantitative work, using rationale examiners actually accept.
Why panels press on sample size
Defending sample size in viva is a specific challenge because sample-size questions have two shapes and it is easy to prepare for only one. The first shape is technical: “why this number?” The second shape is interpretive: “what does the size let you claim, and what does it not?”
Panels press on both. A candidate who answers only the technical question (“I calculated 180 with power analysis”) leaves the interpretive question (“so what can and cannot the study support?”) open. The strong answer covers both in one delivery.
Qualitative research: information power over saturation
For qualitative studies, the concept panels most often accept is information power, framed by Malterud, Siersma and Guassora (2016) in Qualitative Health Research. Information power says the required sample depends on five factors:
The aim of the study. Narrow aims need smaller samples; broad aims need larger.
The specificity of the sample. Highly specific samples (participants with a rare condition, in a defined context) need fewer participants because each carries more information for the aim.
Use of established theory. Studies grounded in existing theoretical frameworks can work with smaller samples because the theory does interpretive work the data would otherwise carry alone.
Quality of dialogue. Well-conducted interviews with rich responses justify smaller samples than interviews that produced thin data.
Analysis strategy. In-depth analysis of a small number of cases justifies fewer participants; cross-case comparison usually needs more.
A qualitative sample-size defence runs through these five factors in a single delivered answer. “My aim was narrow. My sample was specifically drawn from Y. I anchored the analysis in Z framework. The interviews averaged 75 minutes with rich transcripts. My analytic approach is in-depth case comparison. Together these support a sample of 12 participants.”
Saturation is a related concept but does less well as the sole rationale. “I stopped at 12 because I reached saturation” is thinner than the information-power framing, and panels increasingly know this.
Quantitative research: power calculation and effect size
For quantitative studies, the standard defence is a power calculation. Name the expected effect size, the alpha level, the desired power, and the resulting minimum sample. “My primary analysis is a comparison of means with expected effect size 0.5, alpha 0.05, power 0.80, requiring a minimum n of 128. My achieved sample of 214 exceeds this.”
Panels press on three follow-ups:
“Where did the effect size come from?” Cite the source: prior study, meta-analysis, pilot data, or theoretical minimum-important-difference. “Cohen’s conventional medium effect” as the answer reads weaker than a specific citation.
“Is your sample representative?” Address selection bias, response rate, and comparison to the target population. If your sample skews on demographic variables, name the skew and its implication.
“What are the confidence intervals?” Point estimates without confidence intervals are weaker than estimates with them. Have the intervals ready for load-bearing findings.
What to say when the sample size is smaller than ideal
Almost every doctoral study has a sample smaller than ideal. Time, resources, access, and ethics all constrain what is possible. The strong move is to name the constraint honestly, then defend what the actual sample supports.
Three moves:
Name the constraint specifically. “The target sample was 250. I achieved 168 due to lower-than-expected response rate in wave three.”
Recalibrate the claims to fit the actual sample. “The primary analysis is powered at the achieved sample. The subgroup analyses in chapter five are exploratory rather than confirmatory.”
Frame the fuller version as the next study. “A replication with the full target sample would test whether the subgroup effects hold.”
Panels reward this honesty. What they react against is pretending the sample was ideal when it clearly was not.
What examiners actually check in the sample
Beyond the number, panels check three specific things:
The sampling strategy. Convenience, purposive, random, snowball, maximum-variation, theoretical. Each has consequences. Match the strategy to the design and defend it against the alternative.
Inclusion and exclusion criteria. Every choice about who was eligible ruled out someone else. Panels ask about the excluded groups and what claims the exclusions make impossible.
Recruitment and response. How did participants come to be in the study? What proportion of those approached agreed? What are the implications for generalisability?
Have specific answers for each. “38 participants” is a data point. “38 participants recruited through three professional networks, response rate 68% of those approached, with the following inclusion criteria” is a defence.
The specific viva language that lands
Three phrasings consistently work when the panel presses on sample size:
“The sample is scoped to what the design can support.” Signals that you understand the design-sample match rather than just having chosen a number.
“The claims match what the sample allows.” Signals calibration between finding and evidence.
“The next study would extend the sample to test whether X holds.” Signals awareness of the wider research programme.
Trafford and Leshem (2008), across 25 doctoral vivas, found that calibrated language on sample size is one of the specific signals panels rate highly.
What Thesisroom does with this
Thesisroom’s defense practice generates sample-size questions anchored to your methodology chapter. Rehearsing the answer against your own text catches the specific passages panels are likely to press.
Chapter Audit reviews the sample-size rationale in your methodology chapter. If the four-part defence (choice, alternative, reason, consequence) is incomplete, the audit flags it before the viva.
Frequently asked questions
How many qualitative interviews is enough for a PhD?
There is no universal number. Common ranges are 12 to 30 for information-power-based studies, higher for grounded theory approaches. The defence is not the number itself; it is the information-power argument that justifies the number for your specific study.
Is saturation a valid defence in a viva?
It can be, but information power (Malterud, Siersma & Guassora, 2016) is a stronger frame at most contemporary panels. Saturation as the sole rationale reads as thin. Combining saturation with information-power factors reads as considered.
What if my sample size is smaller than my power calculation required?
Name the shortfall honestly. Recalibrate the claims: analyses at the achieved power are confirmatory; underpowered subgroup analyses are exploratory. Frame the fully powered replication as the next study. Panels reward the honesty; they react against pretending the sample was ideal.
Should I have a pilot sample to justify my main sample?
Not always necessary. A pilot strengthens the defence when the effect size or feasibility was unclear at the start. If you did a pilot, cite it in the sample-size rationale. If you did not, the rationale rests on prior literature and theoretical grounds.
What if the panel says my sample is too small?
Do not concede immediately. Run the information-power or power-calculation argument for your specific study, and address the panel’s specific concern. If the concern is about generalisability, name the scope your finding supports. If about statistical power, name what the sample can and cannot support.
Can Thesisroom rehearse sample-size questions with me?
Defense Practice generates sample-size questions anchored to your methodology chapter. Rehearsing the information-power or power-calculation defence against your own text is what makes the answer settled on the day. Chapter Audit reviews whether the defence is complete on the page.
One takeaway before you close this tab
Defending sample size in a viva runs on two moves: an information-power argument for qualitative work or a power calculation for quantitative work, plus calibrated claims that match what the sample supports. When you want the rehearsal against your own chapter, Thesisroom’s defense practice generates the anchored questions.
References
Malterud, K., Siersma, V. D., & Guassora, A. D. (2016). Sample Size in Qualitative Interview Studies: Guided by Information Power. Qualitative Health Research, 26(13), 1753-1760.
Trafford, V., & Leshem, S. (2008). Stepping Stones to Achieving your Doctorate. Open University Press.