What is test-retest reliability?

What is test-retest reliability?

• A. The degree to which a test yields consistent results across multiple administrations under the same conditions.
• B. The variance of scores within a single administration.
• C. The correlation between two different tests.
• D. The accuracy of measurement instruments.

Answer: (A)

Test-retest reliability is about stability over time. If you give the same test to the same people on two separate occasions under similar conditions and the trait being measured hasn’t changed, the scores should be similar. A high degree of similarity is shown by a strong correlation between the two administrations, indicating the test produces consistent results across time. Context helps: this type of reliability matters for traits that are stable. If the interval is too short, people might remember answers; if it's too long, the trait itself might change, which can lower the correlation. The other ideas describe different concepts. Variance within a single administration refers to how spread out the scores are in one testing session. Correlation between two different tests assesses how similarly two different measures relate, not the stability of one test over time. Accuracy of measurement instruments touches on validity or precision rather than consistency across repeated testing.

Test-retest reliability is about stability over time. If you give the same test to the same people on two separate occasions under similar conditions and the trait being measured hasn’t changed, the scores should be similar. A high degree of similarity is shown by a strong correlation between the two administrations, indicating the test produces consistent results across time.

Context helps: this type of reliability matters for traits that are stable. If the interval is too short, people might remember answers; if it's too long, the trait itself might change, which can lower the correlation.

The other ideas describe different concepts. Variance within a single administration refers to how spread out the scores are in one testing session. Correlation between two different tests assesses how similarly two different measures relate, not the stability of one test over time. Accuracy of measurement instruments touches on validity or precision rather than consistency across repeated testing.