National data was obtained from 9-year-old, 13-year-old, 17-year-old, and 26 through 35-year-old populations in order to determine academic achievement in nine subject areas. For each age population, group data was calculated and reported by region, sex, color, parents' educational level, and size and type of community. The application of singular value decomposition of nonsquare matrices to this data is described and its relationship to principal components analysis and its data reduction value is explained. Exploratory analyses are being conducted to determine if the same bases occur across age levels, across time from one assessment to its reassessment, and across subject areas. Emphasis will remain on trying to relate the characteristics of exercises to major differences in performance through the use of orthogonal components. (Author/BJG) Primary type of information provided by report: Procedures (Analysis).