This paper addresses the use of multiple regression in the analysis of reading and math scores. The data presented in the paper were from the Mesa Public Schools in Arizona, and were compiled during the 1975-76 school year. Five criterion factors were selected for investigation: (1) third grade Stanford reading scores, (2) third grade criterion reading scores, (3) third grade criterion math scores, (4) fifth grade Stanford math scores, and (5) sixth grade criterion math scores. The criterion reading and math scores were for the Mesa-developed criterion reading and math tests. Other data was then summarized to form eleven demographic factors that could be used to predict the value of these criterion variables for each of the schools for the 1975-76 school year. Through the use of multiple regression, predicted or expected values were determined for individual schools. These values were then used to compare with the actual scores the school attained. Through the use of multiple regression an individualized predicted value or mean value was determined for each school based on its demographic characteristic. Comparisons of achievement of the school were then made based on its individual characteristics. This method was then contrasted with simple comparisons of a school's achievement with the district mean value and national norm mean. (Author/MV)