In the 1970s, the California Legislature determined that California's curriculum was not well served by national standardized tests and developed its own testing program, the California Assessment Program (CAP). Folklore gives several rationales for this action: urban superintendents wanted to conceal their students' low achievement, which would be revealed with the publication of national norms; consciousness that the California population was more diverse than the U.S. population; or awareness that California's curriculum differed from the national composite used for national tests.
California has had state-wide frameworks to guide district curricula for many years. This story begins with the 1980 addendum to the 1975 framework [California 1982], a small volume that declared “problem solving” to be the umbrella for all of the framework's curricular strands (number, algebra, geometry, statistics, etc.). The task for CAP was to provide the state and districts with information about performance on these strands. It used matrix sampling and item response theory to provide detailed analysis, on the basis of which instructional improvements could be made. CAP was not designed to yield individual student scores; these could continue to come from standardized tests so that parents, teachers, principals, and superintendents could answer the question, “How does this student, this class, this school, or this district stack up in comparison with national norms?”
CAP was designed to provide scores on mathematical topics. It used matrix sampling to make sure that enough kids took items such as “whole number division” or “similar figures” to yield a reliable and valid score.