In a clinical trial with a continuous outcome, the investigators use a paired t-test instead of an unpaired t-test. The primary statistical consequence of correct pairing is:

• A Elimination of type I error in hypothesis testing
• B Reduction of between-subject variability, thereby increasing statistical power ✓
• C Allowance for non-normal distribution of outcomes
• D Control of multiple testing error

Explanation

Pairing subjects (e.g., before-after, twin studies, matched case-control) removes inter-individual variability from the error term, leaving only within-pair (intra-individual) variance. This reduces the standard error of the mean difference and increases statistical power, enabling detection of smaller true effects with a given sample size. Pairing does not eliminate type I error — the alpha level remains the same. For non-normal distributions, Wilcoxon signed-rank (not paired t-test) is used. Multiple testing corrections are addressed with Bonferroni or FDR methods.

Reference: Park's Textbook of Preventive and Social Medicine, 27th ed.

High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP

Written and medically reviewed by the StethoPrep medical team.