ABSTRACT
Reliability measures study the consistency of different aspects of measurement. For instance, when studying interrater reliability, researchers usually obtain the two measurements at the same time i.e., the two observers observe the same participants at the same time. They are seeking to compare the raters to see if they are rating the same incidence in the same or very similar ways. By contrast, in test-retest reliability, researchers measure at two different points in time. This reliability measure seeks to know more about the reliability of the test, not the observers. Suppose a researcher wants to know the reliability of a new test designed to assess the ability to learn college-level math. The researcher might administer the test one week and then re-administer it two weeks later. Because the ability to learn college-level math should not change very much from week to week, the researcher would expect the scores from the two administrations to be consistent.
