Test-Retest Reliability and R2
Keith S. Cox, Zealure C. Holcomb · 2021
To determine test-retest reliability, a test is administered to one group of examinees twice (often with a week or two intervening between the two administrations of the test). This yields two scores for each examinee. Computing a correlation for the relationship between the two sets of scores reveals the stability of the scores over time. A high test-retest correlation indicates that those who scored highly on the first administration tended to score highly on the second administration and vice versa. Typically, if a researcher is interested in test-retest reliability, it is because they expect it to be high, like a correlation of .75 or higher. Correlations of .85 or higher indicate especially strong test-retest reliability.