Power Analysis Calculator

This calculator computes the sample size a study needs using the standard power-analysis formula: baseSampleSize = ((Z-Alpha + Z-Beta) / Effect Size)² (line 18), where Z-Alpha comes purely from Significance Level (α) (line 14) and Z-Beta comes purely from Statistical Power (line 15) — each is a one-input calculation, unaffected by anything else on the page. Statistical Power and Effect Size are the two strongest drivers of Required Sample Size, moving it in opposite directions: raising Power (demanding a stronger detection guarantee) increases the sample size needed, while raising Effect Size (an easier-to-detect difference) decreases it. Test Type changes how that base number becomes Required Sample Size and Total Subjects Needed: Two-sample (Test Type 2) multiplies the per-group requirement by 2 and then doubles it again for the total (lines 25-26), roughly quadrupling total subjects versus a one-sample design at the same settings.

Test Type 3 (Paired) and the One-sample default, despite being offered as separate dropdown options, run through literally identical code (lines 29-30 and 33-34) and always produce the same number. This calculator does not account for unequal group sizes or unequal variances between groups in the two-sample case — it applies a flat ×2 adjustment assuming both groups are the same size with comparable variance, an assumption real unbalanced study designs often violate.