Correct Response: A. Test reliability refers to the degree to which an assessment yields consistent results. A standardized test of English language proficiency for English learners is considered reliable when it produces stable and consistent results. For example, when an English learner's skills are measured on two occasions (with no instruction in between) with the same test, the two sets of scores should be similar. Or, when alternate forms of a test are given to the same English learners, or a matched student population, on two occasions, the scores on the two forms should show a high degree of reliability. The principle described in B is incorrect because it addresses the predictive validity of a test, not the reliability of a test. Neither C nor D describe circumstances or outcomes associated with the reliability of standardized tests of English language proficiency.
