Sample size calculation is one of the most consequential decisions in research design, yet it is frequently treated as an afterthought. An underpowered study cannot detect real effects even when they exist, leading to false negative results and wasted resources. An overpowered study wastes participant time, research funding, and researcher effort on data that provides diminishing returns.
The correct sample size depends on four inputs: the expected effect size, the desired statistical power (typically 80% or 90%), the significance level (α = 0.05 is standard), and the variability in your outcome measure. For most research, you will also need to account for expected attrition or dropout, which means inflating the calculated sample by 10–20%.
Different study designs require different calculation approaches. For comparing two independent group means, use a two-sample t-test power calculation. For proportions (e.g., prevalence studies), use a binomial proportion formula. For correlation studies, use Fisher’s r-to-z transformation. For SEM or factor analysis, rules of thumb suggest 10–20 participants per parameter estimated.
Software options for sample size calculation include G*Power (free, widely cited), nQuery Advisor (commonly required by journals), PASS (comprehensive for clinical trials), and R packages including pwr and pwrss. Your target journal or ethics committee may specify which software is acceptable.
We provide sample size calculation services with full documentation of the formula used, the parameter inputs, the justification for each assumption, and the resulting minimum sample size. Our calculations are formatted as a complete methodology paragraph ready for inclusion in your thesis proposal or ethics application.
