ABSTRACT
This study evolved as an attempt to answer a very specific question about the correlation of trends in tested ability among different data bases: What is the correlation between the mean scores on the NAEP reading test and the mean score on the SAT verbal test for the same cohort of students? The answer to that question was to be pursued using available data from existing data bases-a very significant constraint. The constraint eliminated the option of drawing a sample of 18-year-old SAT takers who had also taken the NAEP reading test at age 13 and then proceeding with a fairly straightforward analysis. Instead, three waves of NAEP reading tests taken by 13-year-olds (in 1971, 1975, and 1980) and three waves of SAT verbal tests (SAT-V) taken by 18-year-olds, 5 years later (in 1976, 1980, and 1985) were examined. There was no reason to assume that any of the 13-year-olds tested by NAEP were also in the sample of SAT takers 5 years later. However, because the samples for each pair of NAEP-SATV means were born in the same year, it was expected that the standardized means would exhibit some degree of correspondence.
