Is Standard Test Scoring a Good Measure of Educational Performance? A Case Study of Public Schools in Connecticut

摘要

To investigate the relationship between students' performance on standard tests and school investment and potential family influence, we collected data for 110 towns or regional school districts in Connecticut and applied standard linear regression model to find out the most significant factors that may affect the test scoring. A chow-test was applied to check if there is a structural difference between the regional school district and the normal school district in each town. Since we used the cross-section data, a test for heteroscedasticity was applied. The result showed that the school investment, in terms of labor and capital inputs, was not important to the students' performance of standard tests, but the household income and parents' educational level seemed positively related to students' performance, and the percentage of non-English home language and percentage of low income families had a negative effect on the scoring.