Power Analysis for Your Dissertation Proposal: A Step-by-Step Guide

April 2, 2026 3 min read By Angel Reyes

Your dissertation proposal probably requires a power analysis to justify your sample size. If you've never done one before, it can feel like a chicken-and-egg problem: you need to know your effect size to calculate your sample size, but you haven't collected data yet.

Don't worry. This guide walks you through it step by step.

What Is a Power Analysis?

A power analysis calculates the minimum number of participants you need to detect a real effect if one exists. It balances four interconnected values:

  1. Effect size — how large the expected difference or relationship is
  2. Alpha level (α) — your threshold for significance (usually 0.05)
  3. Power (1 - β) — the probability of detecting the effect (usually 0.80)
  4. Sample size (n) — the number of participants needed

You set three of these values and solve for the fourth. For your proposal, you'll typically set the effect size, alpha, and power, then solve for sample size.

Step 1: Determine Your Expected Effect Size

This is the part that trips people up. You have three options:

  • Use prior research. Look at similar studies in your literature review. What effect sizes did they find? This is the gold standard.
  • Use Cohen's benchmarks. Small (d = 0.2), medium (d = 0.5), or large (d = 0.8) for group comparisons. See our effect size guide for benchmarks by test type.
  • Use the minimum meaningful effect. What's the smallest effect that would be practically important in your context?

Most dissertation committees prefer that you cite effect sizes from prior studies. If the literature reports d = 0.45, use that.

Step 2: Choose Your Alpha Level and Power

For most dissertations:

  • Alpha = 0.05 (the standard threshold)
  • Power = 0.80 (80% chance of detecting the effect)

Some fields or committees prefer power of 0.90 or 0.95. Check your program's guidelines.

Step 3: Run the Calculation in G*Power

G*Power is free software that handles power analysis for virtually any statistical test. Here's how to use it:

  1. Download G*Power from gpower.hhu.de
  2. Select "A priori" under Type of Power Analysis
  3. Choose your test family (t-tests, F-tests, chi-square, etc.)
  4. Choose your specific test (e.g., independent t-test, one-way ANOVA)
  5. Enter your effect size, alpha, and power
  6. Click Calculate

G*Power will output the minimum sample size per group and total.

Step 4: Write It Up for Your Proposal

Here's a template you can adapt:

"An a priori power analysis was conducted using G*Power 3.1 (Faul et al., 2009) to determine the minimum sample size. With an anticipated medium effect size (d = 0.50), alpha of .05, and power of .80, the analysis indicated a minimum of 64 participants per group (N = 128 total) for an independent samples t-test."

Common Pitfalls

  • Don't just guess "medium." Committees see through this. Ground your effect size in the literature.
  • Account for attrition. If you expect 20% dropout, recruit 20% more than the minimum.
  • Match the test. Your power analysis must use the same statistical test you plan to run. A power analysis for a t-test doesn't apply to an ANOVA.

What If You Can't Get Enough Participants?

If your power analysis says you need 200 participants but you can only access 80, you have options. Consider using a larger effect size threshold, adjusting your research design, or honestly discussing the limitation in your proposal.