Which form evaluates agreement among scorers or across items within a test?

• A. Test-retest reliability
• B. Inter-rater/internal consistency reliability
• C. Sensitivity
• D. Specificity

Answer: (B)

Reliability is about consistency of measurement. The form that focuses on agreement among scorers or across items within a test combines two aspects: inter-rater reliability and internal consistency reliability. Inter-rater reliability asks whether different evaluators score the same behavior in the same way, which is crucial when multiple clinicians or researchers rate a behavior or symptom. Internal consistency reliability examines whether the items on a test all reflect the same underlying construct and thus tend to correlate with one another. Together, these aspects ensure that a test yields stable, coherent scores across raters and items. Statistics like Cohen’s kappa or intra-class correlations capture agreement between raters, while Cronbach’s alpha assesses how well the items work together as a set. By contrast, test-retest reliability measures score stability over time, and sensitivity and specificity relate to diagnostic accuracy rather than agreement or internal consistency.