PHARM/PCS V5 · Procedure 40 Pairwise and Nonparametric
t-Test II: Paired Data
Test whether paired measurements differ, such as before and after on the same subject, with each subject serving as its own control.
What it computes
Tests whether paired observations differ significantly. Enter two groups of equal size; differences are computed automatically.
Inputs and outputs
You provide
- Two groups of equal size, in matching order, such as Before and After
PHARM/PCS reports
- The mean difference and its standard deviation
- The t statistic and its significance
- A plot with one line per subject
When to use it
When the two measurements are naturally paired (before/after on the same subject, left/right, matched littermates) so each subject serves as its own control. Real pairing removes between-subject variability and is far more powerful than treating the data as two independent groups (Procedure 39).
Reading the results
The test works on the differences, so its assumption is that those differences are roughly normal, not the raw values. Only pair when there's a genuine link between the members of each pair; pairing unrelated observations wastes power and misstates the degrees of freedom.
This page describes the procedure. Calculations run inside the licensed PHARM/PCS V5 application. PHARM/PCS is a trademark of MCS Scientific LLC.