| The construct equivalence of the mathematics section of the Scholastic Aptitude Test I (SAT I-M) and the Prueba de Aptitud Academica (PAA-M) was investigated. This study examined whether the psychometric approach known as the rule-space methodology (Tatsuoka & Tatsuoka, 1987), a recently developed statistical technique for identifying and classifying the cognitive strategies used by examinees as they solve test items, is a more effective tool for assessing construct equivalence than traditional psychometric methods. The sample consisted of a group of Spanish-English bilinguals who had taken both tests as part of a field trial conducted by the College Board.; Results of the exploratory and confirmatory factor analyses using the structural equation models suggested that a single common factor explained performance on both tests. These results provide preliminary evidence for the construct equivalence of the two mathematics tests. However, the results obtained using the rule-space methodology seem to lead to a different conclusion: that the parallelism of the SAT I-M and the PAA-M is questionable.; The results underscore the promise of the rule-space approach for guiding test development, and for helping to ensure a focus on test construction that would produce construct equivalence at the level of cognitive strategies. The PAA, which is designed to serve as the Spanish-language version of the SAT I, may need to be revised at the cognitive attribute level to guarantee its parallelism to the SAT I mathematics test. |
