Which reliability measure indicates how consistent the internal items of a test are?

The measure that indicates how consistent the internal items of a test are is Kuder-Richardson Reliability. This measure specifically assesses the consistency of responses across multiple items within a test that are intended to measure the same construct or domain. It is a type of reliability used for tests with dichotomous responses, such as correct/incorrect or yes/no.

The Kuder-Richardson formula calculates a reliability coefficient, providing a numerical value that reflects the degree to which all items on the test correlate with one another. A higher coefficient indicates greater internal consistency, which means that the items reliably measure the same underlying trait or ability.

In contrast, other options address different concepts; for example, test bias refers to the fairness of a test’s content and not its internal consistency. Split-Half Reliability, while related to reliability measures, refers specifically to dividing a test into two halves and comparing the results, rather than assessing the consistency of internal items per se. A confidence interval provides a range of values that estimates where a true score may lie but does not directly measure reliability in terms of item consistency.