PHARM/PCS V5 · Procedure 39 Pairwise and Nonparametric
t-Test I: Grouped Data
Test whether two independent group means differ, with Student's t and a pooled variance, even when the groups differ in size.
What it computes
Tests whether two independent sample means differ significantly using Student's t with pooled variance. Groups may have unequal sizes.
Inputs and outputs
You provide
- The values for each of the two groups, entered as two blocks
PHARM/PCS reports
- The pooled variance
- The t statistic and its significance
- A plot of both groups with their means and confidence intervals
When to use it
For comparing exactly two independent groups on a continuous measure. With three or more groups, don't run repeated t-tests; use one-way ANOVA (Procedure 33) to control the error rate. For the nonparametric alternative when normality is doubtful, see Mann-Whitney (Procedure 45); for naturally paired data, use Procedure 40.
Reading the results
The pooled-variance form assumes comparable variances and roughly normal data; badly unequal spread undermines it. A non-significant result is not proof of equality, since with small samples the test may simply lack power, so weigh the confidence interval on the difference, not just the p-value.
This page describes the procedure. Calculations run inside the licensed PHARM/PCS V5 application. PHARM/PCS is a trademark of MCS Scientific LLC.