A test-retest design is compromised when which issue occurs between tests?

A test-retest design is compromised when which issue occurs between tests?

• A. Random sampling error
• B. Real change in patient function between tests affects reliability
• C. Observer bias
• D. Equipment drift

Answer: (B)

Test-retest reliability depends on the measured trait staying the same between two assessments. If the patient’s function actually changes between tests, the difference in scores reflects a real change in the underlying ability, not inconsistency in the measurement. That real change makes it impossible to interpret the second score as a check of the instrument’s stability, so reliability is compromised. Random sampling error relates to variability in sampling from a population, not the stability of a single individual's score over time. Observer bias would affect scoring consistency across raters or occasions, but the key issue here is the patient's actual change. Equipment drift could introduce measurement error, but the scenario specifically points to changes in the patient’s function between tests as the disruptor of reliability.

Test-retest reliability depends on the measured trait staying the same between two assessments. If the patient’s function actually changes between tests, the difference in scores reflects a real change in the underlying ability, not inconsistency in the measurement. That real change makes it impossible to interpret the second score as a check of the instrument’s stability, so reliability is compromised.

Random sampling error relates to variability in sampling from a population, not the stability of a single individual's score over time. Observer bias would affect scoring consistency across raters or occasions, but the key issue here is the patient's actual change. Equipment drift could introduce measurement error, but the scenario specifically points to changes in the patient’s function between tests as the disruptor of reliability.