An examination of teacher satisfaction using school averages data and LISREL

This study attempted to explain associations among school-level contextual, relational, and teacher satisfaction variables. A large southeastern school district comprised of 149 separate regular school campus' at the elementary, middle, and high school levels provided the setting for this study. School averages data were aggregated by campus for all variables included in the study. School-level contextual variables included measures of student self concept, student performance, student behavior and the degree of impediments to teaching in the school as measured by selected items on a teacher survey. Relational variables included measures of teacher satisfaction, teacher cohesion and teachers' attitudes toward the administration of their schools.

Structural equation modeling, particularly the LISREL statistical software package, was employed as the technique for analysis. The.10 level of statistical significance was chosen for the critical test which determined the degree of correspondence between the model- generated and observed data matrices. The null hypothesis was not rejected at this level. This finding indicated correspondence between the two matrices and suggested that influential paths of association had been successfully estimated.

Results demonstrated that relational variables predict teacher satisfaction more so than other variables. Teacher cohesion and teachers' views of the administration played significant roles in predicting the level of expressed teacher satisfaction. The contextual variables of student behavior and impediments to teaching likewise predicted teacher satisfaction. Student self concept was revealed to be an insignificant predictor of other variables.