How to Use the G*Power Sample Size Justification Prompt to Write a Power Analysis Reviewers Accept

Committees, IRBs, and grant reviewers want to know why you plan to recruit the number of participants you chose. A sample size section that picks a medium effect size without explaining where it came from is a common reason for revisions. The G*Power Sample Size Justification Writer: A Priori Power Analysis Settings by Test Family, Effect Size from Prior Studies or a Smallest Effect of Interest, Attrition Inflation, and a Methods Paragraph prompt maps your planned analysis to the right G*Power settings, derives the effect size from evidence, adds attrition, and writes the justification paragraph for your methods section.

What the prompt produces

1. An analysis mapping to the G*Power test family, statistical test, and a priori power analysis type.
2. An effect size derivation with every step shown, including a small sample bias correction when it applies.
3. A recommended effect size with reasoning, favoring a smallest effect size of interest or a meta analysis over a single small study.
4. The G*Power input panel exactly as you will type it, plus the output fields to copy back.
5. A sensitivity line showing the sample needed for a larger and a smaller plausible effect.
6. Attrition inflation rounded up and balanced across groups.
7. A methods paragraph written for your audience.

How to fill the inputs

StudyDesign describes the design and the primary analysis exactly as planned. The analysis decides which G*Power test applies.

PrimaryOutcome is the main measure and how it is scored.

EffectSizeEvidence holds the numbers you have: means, standard deviations, and sample sizes from a prior study, a meta analysis estimate, or a smallest effect size of interest with a reason.

ErrorRates sets alpha, power, tails, and any correction for multiple primary outcomes.

Attrition is the expected loss and why you expect it.

Audience tells the prompt who will read the justification, such as an IRB, a thesis committee, or grant reviewers.

Reading the example output

The example plans a two group trial comparing a study skills workshop to a control session, with final exam score as the outcome:

• The mapping picks the independent means test in the t tests family, matching the planned independent samples t test.
• The effect size from the prior study is computed step by step. The pooled standard deviation and Cohen's d come out near 0.5, and the Hedges correction lowers it slightly.
• The recommendation uses the department's smallest meaningful difference, a 3 point gain, which corresponds to d of 0.40.
• The input panel lists every field, and the expected output is labeled for the user to confirm by running G*Power.
• The sensitivity line shows the difference between powering for d of 0.50 and d of 0.40, which changes the target by 72 students.
• Attrition of 15 percent raises the recruitment target and keeps the groups equal.
• The methods paragraph names the test, every input, the effect size source, and the final numbers.

Tips for better results

• Paste the actual statistics from the source study rather than a reported effect size label.
• Decide your smallest effect size of interest with your advisor or team before running anything.
• Run G*Power yourself and paste the output so the paragraph uses confirmed numbers.
• Base attrition on records from similar studies or courses, not a round guess.
• Keep the G*Power screenshot in your protocol files.

Mistakes to avoid

• Do not justify an effect size with small, medium, or large labels alone.
• Do not trust a single pilot study's effect size without adjustment or a second source.
• Do not use G*Power for designs it cannot model well, such as complex multilevel models. Use simulation instead.
• Do not forget to inflate for attrition before setting the recruitment target.

Who it is for

Graduate students writing proposals, faculty preparing grant applications, IRB coordinators reviewing protocols, researchers preregistering studies, and methods consultants who write sample size sections for others.

Related PromptDig links

Open the G*Power Sample Size Justification Writer: A Priori Power Analysis Settings by Test Family, Effect Size from Prior Studies or a Smallest Effect of Interest, Attrition Inflation, and a Methods Paragraph prompt and describe your design. For more research prompts, Browse more prompts. If you have a research methods prompt that works, Share a prompt.