| This study examined the relationships between salaries of 1000 Texas school superintendents and selected indicators of student achievement, student demographics, and per student expenditures. Regression analyses were run with the dependent variable, superintendent salaries, and the following independent and intervening variables: student achievement as reflected by Academic Excellence Indicator System categories, attendance rates, disciplinary placement percentages, racial/ethnic categories, economically disadvantaged percentages, total student population, total per student expenditures, and per student expenditures for central office administration. A series of complete and reduced linear models, regression equations, and F-tests of those models were run. The regression models allowed for the prediction of superintendents' salaries.; The results fell into three categories. The complete model of all independent and intervening variables was strongly predictive of the dependent variable, superintendents' salaries, with an R2 of .686. One reduced model, using total student population as an independent variable, was also strongly predictive with an R2 of.572. Three reduced models were moderately predictive: racial/ethnic categories (R2 = .186), per student central office expenditures (R2 = .109), and total per student expenditures (R2 = .064). The remaining models were not predictive of superintendents' salaries.; This study provided useful information that can be utilized in demystifying the salary determination processes for superintendents in Texas and developing a model whereby sound educational standards are utilized in making salary decisions. A research based understanding of how superintendent salaries relate to student demographics, student achievement, and per student expenditures can empower decision makers in their task of paying leaders. |
