The t test for two related samples is appropriate for evaluating a treatment effect by comparing the measures of the dependent variable before and after the treatment in a repeated measures design experiment. Shaughnessy, Zechmeister, and Zechmeister (2012) explained that “in a repeated measures design experiment, researchers manipulate an independent variable to compare measures of participants’ behavior in two or more conditions” (p. 228). For example, in a research study in comparison of job-lost men's self-esteem level before and after a treatment for depression, one sample is used for a repeated-measures t test thus the same subjects are tested with the same procedure to measure the dependent variable before and after the treatment. Using the experimental data, the mean difference of the two measurements can be analyzed to determine if the treatment has significantly changed the subjects’ self-esteem level.
The repeated-measures t test is the most appropriate in this case because “the primary advantage of a repeated-measures design is that it reduces or eliminates problems caused by individual differences” (Gravetter & Wallnau , 2013, p. 367). These individual differences such as age, personality, IQ, and personal experience can influence the self-esteem assessment test scores and affect the outcome of the hypothesis test. When using repeated measures, researchers can ensure that the participants in one treatment make no difference from those in another treatment thus it is not likely that any difference in personal factors could interfere the hypothesis test. Also, a repeated-measures test requires fewer subjects than that in an independent-measures test thus the experiment is easier and more convenient to conduct.
Gravetter, F. J. & Wallnau, L. B. (2013). Statistics for the behavioral sciences (9th ed.). Belmont, CA: Wadsworth.
Shaughnessy, J. J., Zechmeister, E. B., & Zechmeister, J. S. (2012). Research methods in psychology (9th ed.). New York, NY: McGraw-Hill.