Appropriate Statistical Methodology to Compare pre- and post-study Results




Sample Paper Excerpt
Two co-authors, both psychologists, have submitted a manuscript describing their study that examined the superiority of a newly developed instrument to measure a general self-efficacy of a person. The study was designed with 20 healthy volunteers being administered the standard Self-efficacy Outcome Study (SOS) mental health scale soon after entering the study. After 3 months they were administered the newly developed instrument to measure the response. The data consisted of pre (SOS scale) and post (new instrument) study results on 20 subjects. The authors decided to test whether the mean post-study result is better than the mean pre-study result using a 'paired' t-test. A reviewer comments that the researcher should be using a standard 'pooled' t-test.

Which is the more appropriate statistics test?
Paired t-test. Pooled t-test.

Proceed to the


Finished?