Which dimension refers to the consistency and stability of scores across administrations, tasks, or raters?

Which dimension refers to the consistency and stability of scores across administrations, tasks, or raters?

• A. Validity
• B. Reliability
• C. Transfer & Generalization
• D. Consequences

Answer: (B)

Reliability is about consistency and stability in measurement. It asks whether scores would be similar if the same person took the test again on another occasion, if different but equivalent tasks yield similar results, or if different raters score responses similarly. When a test is reliable, observed score differences reflect real changes in ability rather than random error. For example, a language assessment with strong test-retest reliability should produce similar scores across administrations under comparable conditions, and an essay rubric with high inter-rater reliability would yield comparable scores from different graders. This focus on consistency distinguishes reliability from validity, which is about whether the test measures what it’s intended to measure, and from transfer/generalization or consequences, which deal with applicability across contexts and the impact of the test on decisions.

Reliability is about consistency and stability in measurement. It asks whether scores would be similar if the same person took the test again on another occasion, if different but equivalent tasks yield similar results, or if different raters score responses similarly. When a test is reliable, observed score differences reflect real changes in ability rather than random error. For example, a language assessment with strong test-retest reliability should produce similar scores across administrations under comparable conditions, and an essay rubric with high inter-rater reliability would yield comparable scores from different graders. This focus on consistency distinguishes reliability from validity, which is about whether the test measures what it’s intended to measure, and from transfer/generalization or consequences, which deal with applicability across contexts and the impact of the test on decisions.