Sample size for repeated-measures ANOVA
Effect size is the difference in means between three observations of an outcome
In order to run an a priori sample size calculation for repeated-measures ANOVA, researcheres will need to seek out evidence that provides the means and standard deviations of the outcome at the three different observations. The absolute differences between these three mean values and their respective variances constitutes an evidence-based measure of effect size. Find articles in the literature that are conceptually or theoretically similar to the study of interest or use similar outcomes and use those values in the sample size calculation for repeated-measures ANOVA.
A good rule of thumb is to overestimate the variance of the effect size. Researchers do this because it forces them to have to collect more observations of the outcome, which in turn leads to more precise and accurate measures of effect with repeated-measures ANOVA.
For example, let's say that researchers find quality evidence that people in the treatment group have an average pain score of 7.1 with a standard deviation of 1.6 at baseline, an average pain score of 4.3 with a standard deviation of 1.1, and a 6-month follow-up average score of 4.1 with a standard deviation of 1.4. Researchers could enter these values into G*Power and know exactly how many observations of the outcome they would need to collect to detect the hypothesized treatment effect.
A good rule of thumb is to overestimate the variance of the effect size. Researchers do this because it forces them to have to collect more observations of the outcome, which in turn leads to more precise and accurate measures of effect with repeated-measures ANOVA.
For example, let's say that researchers find quality evidence that people in the treatment group have an average pain score of 7.1 with a standard deviation of 1.6 at baseline, an average pain score of 4.3 with a standard deviation of 1.1, and a 6-month follow-up average score of 4.1 with a standard deviation of 1.4. Researchers could enter these values into G*Power and know exactly how many observations of the outcome they would need to collect to detect the hypothesized treatment effect.
The steps for calculating sample size for a repeated-measures ANOVA in G*Power
1. Start up G*Power.
2. Under the Test family drop-down menu, select F tests.
3. Under the Statistical test drop-down menu, select ANOVA: Repeated measures, within factors.
4. Under the Type of power analysis drop-down menu, select A priori: Compute required sample size - given alpha, power, and effect size.
5. Click the Determine button.
6. Click on the Direct marker to highlight the menu.
7. In the Partial eta-squared box, enter one of the following values:
Enter ".01" if researchers believe there will be a small treatment effect.
Enter ".03" if researchers believe there will be a moderate treatment effect.
Enter ".05" if researchers believe there will be a large treatment effect.
8. Click Calculate.
9. Click Calculate and transfer to main window.
10. Enter .80 into the Power (1-beta err prob) box, unless researchers want to change the power according to the current empirical or clinical context.
11. In the Number of groups box, enter "1"
12. In the Number of measurements box, enter "3"
13. Click Calculate.
2. Under the Test family drop-down menu, select F tests.
3. Under the Statistical test drop-down menu, select ANOVA: Repeated measures, within factors.
4. Under the Type of power analysis drop-down menu, select A priori: Compute required sample size - given alpha, power, and effect size.
5. Click the Determine button.
6. Click on the Direct marker to highlight the menu.
7. In the Partial eta-squared box, enter one of the following values:
Enter ".01" if researchers believe there will be a small treatment effect.
Enter ".03" if researchers believe there will be a moderate treatment effect.
Enter ".05" if researchers believe there will be a large treatment effect.
8. Click Calculate.
9. Click Calculate and transfer to main window.
10. Enter .80 into the Power (1-beta err prob) box, unless researchers want to change the power according to the current empirical or clinical context.
11. In the Number of groups box, enter "1"
12. In the Number of measurements box, enter "3"
13. Click Calculate.
Based on the values and steps above, for one group being measured across three observations, an alpha of .05, and a power of .80, and small treatment effect of .01, researchers would need to collect 161 observations to detect a significant treatment effect.
Click on the Statistics button to continue.
Statistician For Hire
DO YOU NEED TO HIRE A STATISTICIAN?
Eric Heidel, Ph.D. will provide statistical consulting for your research study at $100/hour. Secure checkout is available with PayPal, Stripe, Venmo, and Zelle.
- Statistical Analysis
- Sample Size Calculations
- Diagnostic Testing and Epidemiological Calculations
- Psychometrics