Which reliability type assesses score stability over time?

Which reliability type assesses score stability over time?

• A. Test-retest
• B. Parallel forms
• C. Internal consistency
• D. Inter-rater reliability

Answer: (A)

Score stability over time is evaluated by test-retest reliability. The idea is to administer the same test to the same individuals on two or more occasions and examine how closely the resulting scores align. If the underlying trait hasn’t changed and the measurement is stable, a high correlation between the two administrations indicates strong reliability. The chosen time interval matters: too short can inflate stability due to memory or practice effects, while too long can allow real changes in the trait to occur, lowering the correlation. In contrast, parallel forms reliability checks consistency across different but equivalent versions of a test given around the same time, focusing on whether the different forms measure the same construct similarly. Internal consistency looks at how well items on a single test hang together at one time point, typically within a single administration. Inter-rater reliability assesses agreement among different evaluators scoring the same performance, not how stable scores are over time.

Score stability over time is evaluated by test-retest reliability. The idea is to administer the same test to the same individuals on two or more occasions and examine how closely the resulting scores align. If the underlying trait hasn’t changed and the measurement is stable, a high correlation between the two administrations indicates strong reliability. The chosen time interval matters: too short can inflate stability due to memory or practice effects, while too long can allow real changes in the trait to occur, lowering the correlation.

In contrast, parallel forms reliability checks consistency across different but equivalent versions of a test given around the same time, focusing on whether the different forms measure the same construct similarly. Internal consistency looks at how well items on a single test hang together at one time point, typically within a single administration. Inter-rater reliability assesses agreement among different evaluators scoring the same performance, not how stable scores are over time.