Effective schools correlates and achievement gain : a relational study

This study was conducted to investigate the relationship between school effectiveness correlates identified in prior research and achievement gain in grades 4-6 in the schools of a medium sized southeastern city. These schools provided a reasonably representative sample and the achievement test records had student and school identification numbers making it possible to gather longitudinal achievement data.

The school sample was composed of all 23 schools in the system which included grades 4, 5 and 6. The Least Squares Means (LSMEANS) method under the General Linear Model Procedure (PROG GLM) of the Statistical Analysis System (SAS) was used to compute a gain estimate in total reading, total math, and total language for each school based on performance of approximately 1,000 students each in the 1985 and 1986 sixth-grade classes on the California and Stanford Achievement Tests which had been administered under the state testing program. All students for whom third, fourth, fifth and sixth grade test records were available were included in gain computation for a school if records indicated attendance in that school during fourth, fifth and sixth grades.

The Connecticut School Effective Questionnaire (CSEQ) was completed by 335 full-time faculty members of the schools and a school mean for each of the seven correlates was computed. These school means were used as independent variables in regression analysis with school mean gain estimates as the dependent variable. Each subject (reading, arithmetic and language) was analyzed separately. None of the independent variables could meet the significance requirement for entry into the regression equation under the Statistical Program for the Social Sciences (SPSSX).

The seven correlates measured by the CSEQ were not related to school effectiveness (achievement) gain in this study.