The relationship of elementary school size to the 1996 Stanford Achievement Test scores of fifth-grade students

The purpose of this study was to examine the effect of school size on fifth grade student Stanford Achievement Test (SAT) scores in larger schools than past studies had considered. This study measured the interactions and main effects of the following independent variables: School size; student ethnicity; and student socioeconomic status (SES) as defined by student lunch status. The 1996 SAT raw math application/math computation scores and raw reading comprehension scores of fifth grade students provided the dependent variables.;The sample of students was randomly selected from three categories of public schools in a large urban district: small (less than 685; ;School size was found to interact with ethnicity such that at large schools, Black students scored higher in math and reading than Black students scored at small and medium schools. In large schools only, there was no significant difference among the math and reading scores of Black students, Hispanic students, and White students. When all student groups were combined, no main effect was found for school size.;The main effect of ethnicity was significant. In reading, White students scored higher than Black and Hispanic students. In math, White students scored higher than Black students and comparable to Hispanic students. Hispanic students scored higher than Black students in math and reading.;There was a main effect for lunch status; paid lunch students scored higher than free lunch students. Lunch status did not interact with size or ethnicity. Nor was there a three-way interaction found among size, lunch status, and ethnicity.