This study analyzed the gap of student's scores of the college entrance exam(that is, Korean SAT)
between municipals, and discussed the issues that could be addressed when test scores are used to
evaluate school quality. It has investigated how school mean scores of Korean SAT changed after
controlling the effect of regional factors, such as the self-reliance ratio of municipal finance, the ratio
of poverty population, and private tutoring. The regional information was matched with the four years
dataset of Korean SAT, and three level hierarchical model method was applied.
In result, about 25.2% to 32.1% of total variance of student achievement was found at the school
level. And the 47.2% to 54.4% of such variance was explained by school and region variables. This
shows that between-school difference of student's achievement was positively associated with the
educational conditions of school and region. But, the growth rate of school mean score were not
related with school and regional factors. These results indicate that the difference of school mean
scores could not be regarded as school's educational competency, and that student backgrounds and
regional conditions should be considered if we intend to use students' test scores in order to evaluate
school performance.