"How many respondents do I need?" is one of the first questions every scholar asks — and one of the easiest to get wrong. There is no universal magic number; sample size follows from your design, your analysis, and the effect you expect to find.
What actually drives sample size
- Population size — for surveys of a finite, known group.
- Confidence level and margin of error — usually 95% confidence, ±5%.
- Expected effect size — smaller effects need larger samples to detect.
- Statistical power — commonly set at 0.80, the chance of detecting a real effect.
- Your analysis — regression, ANOVA and SEM each carry their own minimums.
Rules of thumb (use with care)
You'll hear heuristics — 10 cases per variable for regression, 200+ for CB-SEM, the 'ten-times' rule for PLS-SEM. They're useful sanity checks, but they don't replace a proper power calculation grounded in your expected effect.
For a large population at 95% confidence and ±5% margin, roughly 384 completed responses is a common target. Adjust upward for low response rates, and downward only for genuinely small, finite populations.
Calculate it, then defend it
Use our free sample size calculator to get a defensible number, then read how much sample size is enough to justify it in your methodology chapter. For power analysis tied to your specific model, sampling mentoring walks you through it.
Frequently asked
Is a bigger sample always better?+
Not necessarily. Beyond the size your analysis needs, extra data adds cost and effort for little gain — and can make trivial effects look 'significant'. Aim for adequately powered, not simply large.
What sample size do I need for SEM?+
It depends on the approach. CB-SEM (AMOS) generally wants 200+ and reasonably normal data; PLS-SEM (SmartPLS) tolerates smaller, non-normal samples. Both are better set by a power analysis than a rule of thumb.
phdguide's mentors are senior academics, former supervisors, statisticians and publication specialists with 25+ years of combined experience guiding MBA, MPhil and PhD scholars from topic to viva.
Ethical, compliant guidance: We provide academic support, mentoring, analysis, editing and structuring — not authorship. Your work stays compliant with university policies.