StatReasonSample Size & Statistical Power
These tools are for planning, used before data collection, not for analyzing data you already have. There is no generic sample size that fits every study: every engine here requires a specific design, a target precision or effect, an alpha level, and a target power or confidence level, because those choices are what actually determine how many observations you need.
Every planner shows a sensitivity view and rounds its recommended sample size up to a whole number, but the plausibility of your assumed effect size, event rate, or standard deviation, and whether your actual recruitment can reach the recommended sample, are left for you to judge.
Engines in this category
- Sample Size for a Proportion
Plan the sample size needed to estimate a proportion to a target precision.
- Sample Size for a Mean
Plan the sample size needed to estimate a mean to a target precision.
- Sample Size for Two Means
Plan an independent two-group mean comparison before collecting data.
- Minimum Detectable Effect Calculator
Find the smallest effect a planned sample size can reliably detect.
- Sample Size Calculator
Route to the sample-size planner built for your specific design.
- Sample Size for Paired Means
Plan a matched-pairs study using the variability of the differences.
- Sample Size for Two Proportions
Plan an independent two-proportion comparison before collecting data.
- Sample Size for Correlation
Plan detection or precision for a target Pearson correlation.
- Sample Size for One-Way ANOVA
Plan a balanced k-group ANOVA using the exact noncentral F relationship.
- Power Calculator for t-Test
Find the achieved or prospective power for a one-sample, paired, or independent design.
- Power Calculator for Proportions
Find the achieved power for a one- or two-proportion design.
- Power Calculator for ANOVA
Find the achieved power for a balanced one-way ANOVA from Cohen's f and n.