Diagnostic Assessment in Timss-r: Comparison of Eighth Graders’ Mathematics Knowledge States in the United States, Japan, and Israel

The study combines between-countries and within-country comparisons of performance on the 1999 TIMSS-R mathematics test using a probabilistic model for cognitive analysis known as "rule space" (RS) (Tatsuoka, 1983, 1995, in press). Samples of 8th graders from three countries – the United States, Japan, and Israel – were compared with respect to mean probabilities of mastering attributes of content, skills, and cognitive processes underlying their performance on the test. Also compared were the proportions of students from the three samples in each of eight hierarchically ordered clusters of knowledge states. The same analyses were employed for an intra-country comparison on the Israeli sample between the two culturally diverse populations of Jews and Arabs, both of whom study according to the same intended curriculum but in separate schools. The results of the betweencountries comparisons indicated the superiority of the Japanese sample in mathematics knowledge and thinking skills. The comparisons of Jewish and Arab students in Israel indicated significant differences in favor of the Jewish group on all attributes. The results are discussed in light of prevalent instruction – learning – assessment cultures in the respective countries and sectors. Also discussed is the utility of RS methodology for large-scale diagnostic assessment targeted at betweencountries and within-country comparisons.