In testing the difference between the means of two normally distributed populations, the number of degrees of freedom associated with the unequal-variances t test statistic usually results in a non-integer number. In this situation, what is the best recommendation in order to proceed?
A) You should round it up to the nearest integer.
B) You should round it down to the nearest integer.
C) You should change the sample sizes until the number of degrees of freedom becomes an integer.
D) You should assume that the population variances are equal, and then use df =.

Question

Quizplus · Accepted Answer

When using the unequal-variances t-test (Welch's t-test), the degrees of freedom are often non-integer. The best practice is to round down to the nearest integer to ensure a more conservative estimate, which reduces the risk of Type I error.

In Testing the Difference Between the Means of Two Normally