Comparing the achievement of eighth-grade boys and girls on norm referenced and performance tests in language arts, reading and mathematics

In this study, the hypothesis that the performance (standards) test type would be a better match for girls than the norm referenced (measurement) test type is examined. With assessment becoming more high stakes and with an increased use of the performance test type, this research appears timely.;In addition to cultural sex differences, the research literature supports the existence of sex differences in brain physiology, laterality, cognitive networking and specific skill strengths. Literature also supports distinctions between test types that call for different skills in respondents. Thus, this study is designed to examine sex and sex by test type differences on exemplars of norm referenced and performance based tests.;Results on the Metropolitan Achievement Test (norm referenced) and the New Standards Reference Exam (performance) are used. To minimize confounding variables, an eighth grade homogeneous suburban population is chosen. Test results are analyzed using ANCOVA for between test differences and t-tests for within test differences.;Scores on the Otis Lennon School Ability Test are used as covariates for four separate ANCOVA. For each analysis, independent variables are sex (boy, girl) and test type (MAT, NSRE). The dependent variables are mean scores on the total test and the subject areas of mathematics, language arts/writing and reading. The data set of 230 is comprised of 118 boys and 112 girls.;Findings support significant sex and sex by test type interactions within a 95% confidence interval. Girls outscore boys on both tests. In same sex comparisons, girls score higher on the NSRE and boys score higher on the MAT. The sex gap widens considerably on the NSRE.;Sex differences within and between tests differ somewhat. On the MAT, girls score significantly better in all areas except mathematics. On the NSRE, girls significantly outscore boys in all areas.;Results indicate the existence of sex differences within and between these test types. The hypothesis that the performance test type is a better match for girls is supported. However, since the sex gap widens with the performance test type, further research regarding sex and other group differences is recommended.