An Investigation of the Test/Retest Stability of the Stanford Achievement Test

Abstract

This study investigated the stability of three subtests of the Stanford Achievement Test. It further analyzed reliability and validity assessment of standardized tests in general, with particular emphasis on their measurement of reading ability. It clarified the position of stability assessment within a total evaluation framework. The subjects of the study were 39 fourth grade students from an urban upstate New York Catholic school district. The Reading Comprehension, Word Study Skills and Vocabulary subtests of the Stanford Achievement Test, Intermediate Level l, Form E were administered to the subjects by their regular classroom teachers during the third week of May. Nine days later the same tests (same form) were re-administered under similar conditions. The raw scores for each test were then paired and correlated for analysis utilizing a Pearson Product Moment procedure. Results of the study indicated that all three subtests were stable in their measurement of student performance over time. Positive, significant relationships existed between test/retest scores for each test. Student completion rate was found to have a major effect on the correlation index of the Reading Comprehension subtest.