The Sample Size & Power Analysis Workbook

64 G*Power Walk-Throughs, Effect Sizes, and the Sample-Size Justification Paragraph Your Committee Demands

"How many participants do I need?" Most students guess: a round number, the figure a friend used, whatever the journals seem to expect. Then a committee member, an ethics board, or a grant reviewer asks where the number came from. A sample size is not a guess. It is a number you derive from four linked parameters and then justify in writing.

The Sample Size & Power Analysis Workbook is not a statistics textbook. It is the working surface for doing the calculation: 64 fill-in templates across thirteen chapters. One fictional trial is worked from the first estimate of an effect size to the finished justification paragraph, a second is worked through correlation and regression, and a third shows how to justify a qualitative sample without a power analysis.

Who this workbook is for: Master's and doctoral students writing a quantitative or mixed-methods proposal, researchers in health, nursing, psychology, education and the social sciences, and supervisors who want a student's sample-size reasoning laid out step by step.

What's Inside

✓

The Right Kind of Power Analysis

A priori, sensitivity and compromise analyses, the inputs each test needs, and why "observed power" is not one of them.

✓

A Power Sheet for Each Common Test

t-tests, ANOVA, ANCOVA, correlation, multiple and logistic regression, chi-square and proportions, each with its inputs, its menu path and a worked result.

✓

Attrition, Recruitment & Qualitative Samples

Inflate your number the right way, work back to how many people you must screen, check you can actually recruit them week by week, and justify a qualitative sample by saturation and information power.

Core Checklists & Templates Included

64 fill-in templates across thirteen chapters, most with a worked example and every one with a supervisor sign-off line
Effect sizes estimated from evidence, a smallest effect worth detecting, and a reported sensitivity range
The designs G*Power cannot size — clustered data, mediation, SEM, equivalence — and where to take them instead
Effect-size conventions, conversion formulas, and the menu path for each test, gathered in the appendices
The sample-size paragraph written out for your proposal, your ethics application, and your grant

G*Power is free software from Heinrich-Heine-Universität Düsseldorf; this workbook is independent of its developers. The worked studies are invented to demonstrate the method, so check every figure in current software for your own study.