Ontario third-grade mathematics assessment results were analyzed on three knowledge/skill dimensions and five content strands. Seventy percent of students had the same subscores on the dimensions (flat profile); 30 percent differed on only one dimension (contoured profile); 69 percent had flat or contoured content profiles. Profiles were used to…

Data from 127 urban and 39 rural elementary schools were used to examine the utility of Canadian census data for understanding differences in school-level academic achievement, and whether the method of constructing demographic variables affected their relationship to achievement. Indicators obtained with the enrollment method correlated more…

Reports results of a comparison of correlational and multidimensional scaling techniques as tools for analyzing the perceptions of three groups of mathematics teachers (at high school, vocational school, and university levels) of the relative importance of curricular objectives. Discusses the relationship of the two analyses and the implications…

Analysis of response strategies employed by 261 undergraduate chemistry students when answering multiple-choice and stem-equivalent constructed-response questions revealed no significant differences in types of solution strategies or types of errors across test format. However, analysis of student oral reports revealed a higher frequency of…

Illustrates the use of forced classification for assessing student solutions to ill-structured problems. The analysis, applied to discussions by children of a social problem, consists of content categorization; quality ratings within content categories; and categorical discriminant analysis, with student grade and judged quality as discriminating…