An Examination of the Relationship Between School Scores Derived From Commercial Achievement Tests and Those From Statewide Assessment

An apparent discrepancy between building level scores in basic skills produced by Pennsylvania's state assessment program (EQA) and building summary scores, generally a grade equivalent, provided by commercial standardized achievement tests is investigated. The impetus for the study came from occasional reports by school administrators that their school level grade equivalent suggested above average performance while the state assessment percentile rank seemed low by comparison. Achievement test scores at the building level, obtained from approximately 302 school districts, were merged with state assessment scores in reading, writing skills and mathematics for grades 5, 8, and 11. Correlations in the .70 to .85 range revealed a similar rank ordering of schools by commercial and state assessment tests. Rather than average building performance being equivalent to the grade-month of testing, as believed by many administrators, the distribution of building mean scores for several commercial tests indicated that "average" corresponded to a grade equivalent of six months to a year beyond that point. Commercial achievement test scores examined included those from the California Achievement Tests, Comprehensive Tests of Basic Skills, Iowa Tests of Basic Skills, Metropolitan Achievement Tests, Science Research Associate Assessment Survey and the Stanford Achievement Test. (Author/RL)