A test yields consistent results across administrations but does not measure the intended construct.

A test yields consistent results across administrations but does not measure the intended construct.

• A. A test yields consistent results across administrations but does not measure the intended construct.
• B. A test is valid but not reliable.
• C. A test is neither reliable nor valid.
• D. A test is perfectly reliable and perfectly valid.

Answer: (A)

Reliability means the test gives consistent results across administrations, while validity means it actually measures the construct it’s intended to measure. If a test yields the same results on different occasions but those results don’t reflect the intended construct, it is reliable but not valid. That’s why the scenario describes a test that is consistent (reliable) yet fails to measure what it’s supposed to measure (not valid). For example, a thermometer that always reads 2 degrees high is reliable (consistent) but not valid for measuring true body temperature. The other options describe situations that don’t match the given scenario: valid but not reliable would show consistency issues; neither reliable nor valid would fail to show reliability; perfectly reliable and perfectly valid would meet both criteria, which the scenario explicitly says it does not.

Reliability means the test gives consistent results across administrations, while validity means it actually measures the construct it’s intended to measure. If a test yields the same results on different occasions but those results don’t reflect the intended construct, it is reliable but not valid. That’s why the scenario describes a test that is consistent (reliable) yet fails to measure what it’s supposed to measure (not valid).

For example, a thermometer that always reads 2 degrees high is reliable (consistent) but not valid for measuring true body temperature. The other options describe situations that don’t match the given scenario: valid but not reliable would show consistency issues; neither reliable nor valid would fail to show reliability; perfectly reliable and perfectly valid would meet both criteria, which the scenario explicitly says it does not.