When participants in a study are measured, given a treatment, and then measured again (called a repeated measures design), the score on the second measures will be very similar to the first because the measurements come from the same person. Experimental groups that are established this way are said to be dependent groups. Dependent groups can also be formed by matching subjects on variables the researcher thinks are important to the outcome of a study. When a researcher is interested in testing the difference between the means of two dependent groups, then the t-test for Dependent Groups is appropriate. This test is also known as the t-test for correlated data and the t-test for paired data. The difference between the t-test for dependent groups and the t-test for dependent groups is a matter of how t is computed. There is no difference in how t is reported and interpreted.
t-Test for Dependent Groups
-The t-test for dependent groups compares the means of two variables for a single group (or two functionally equivalent groups as with matching).
-The test computes the differences between scores on the two variables (e.g., pretest and posttest scores) for each case and tests whether the average differs from zero.
Last Modified: 02/12/2019