Two-Sample t Test Assuming Unequal Variances 1. Refer to Example 10.25. If the ratio of the two sample variances had been so large that you could not assume equal variances (we use "greater than 3" as a rule of thumb), you should select Stat Basic Statistics 2-Sample t, but DO NOT check the box marked "Assume Equal Variances." If we use this test on the data from Example 10.25, the following output results (Figure 10.27).

2. You will see slight differences in the degrees of freedom; there is no "Pooled StDev" listed, but the conclusions did not change.

Session Two-Sample T-Test and CI: Teacher A, Teacher B Teacher & 1.65 9.91 Teacher B 1.78 0.00 3.5 Difference (Teacher A) (Teacher B -0.15 Estimate for difference: 95% CI for differences T-Test of difference (-9.96, 9.65) 0 (va not): I-Value -0.03 -Value - 0.374