Define reliability in the context of psychological measurement.

Define reliability in the context of psychological measurement.

• A. The consistency or repeatability of test scores across occasions, raters, items, or forms.
• B. The extent to which a test measures what it intends to measure.
• C. The degree to which a test produces higher scores for certain groups.
• D. The speed at which a test is completed.

Answer: (D)

Reliability is about consistency and stability in measurement. A reliable test yields similar scores if the same person takes it again on a different occasion (assuming the trait hasn’t changed), and different raters would score the same responses in the same way (inter-rater reliability). It also reflects coherence among the test items, meaning the items collectively measure the same underlying construct (internal consistency). In short, reliability is about precision and the reduction of random error in the observed scores. The statement that best captures this idea is the one describing consistency of scores across occasions, raters, items, or forms. The other notions—whether the test measures what it intends to measure (validity), whether the test favors certain groups (bias), or how quickly the test is completed (speed/efficiency)—do not define reliability.

Reliability is about consistency and stability in measurement. A reliable test yields similar scores if the same person takes it again on a different occasion (assuming the trait hasn’t changed), and different raters would score the same responses in the same way (inter-rater reliability). It also reflects coherence among the test items, meaning the items collectively measure the same underlying construct (internal consistency). In short, reliability is about precision and the reduction of random error in the observed scores.

The statement that best captures this idea is the one describing consistency of scores across occasions, raters, items, or forms. The other notions—whether the test measures what it intends to measure (validity), whether the test favors certain groups (bias), or how quickly the test is completed (speed/efficiency)—do not define reliability.