The crucial issue that led to this dissertation is whether money does matter in students` standardized test achievement. The objective of this dissertation is to find out if there exist any statistical significance between the MEAP test results of met/exceeded standard and per pupil expenditure, per pupil/teacher ratio (class size), teacher quality, adults educational attainment level, adults income level and minorities in Michigan Public Schools at elementary, middle and high school levels. Since Coleman`s report (1960), educational research on standardized test achievement had been conducted with ideological predilection. This research came to be conducted as purely empirical study, neutral to any of the political ideological lines in education, to determine variables that relate to the MEAP test results of met/exceeded standard.
The research was based on Michigan Public Schools MEAP test results of met/exceeded standard at elementary, middle and high school levels. The force entry multiple regression method was applied based on the data given to determine which of the explanatory variables explain the MEAP test results of met/exceeded standard. In addition, lagged variables were use to determine how time factor plays for some variables to be used to explain the MEAP results of met/exceeded standard in Michigan public elementary, middle, and high school levels.
By using the force entry multiple regression and forced entry multiple regression with lagged variables, the descriptive statistics were use to analyze the means and standard deviations to demonstrate performances at each school level in Michigan public school districts. In the overall analyses of the results to determine the relation between the outcome and explanatory variables, all the estimates used included the beta coefficients, Durbin-Watson, Adjusted R square, the ANOVA, overall significance level, the t statistics and the significance of the coefficients.
If a particular school level overall model was not significant, other statistical test estimates were looked at and further diagnostic tests were conducted to determine if none of the assumptions was violated. In this perspective, Variance Inflation Factor (VIF) estimate was used to check collinearity problems, the Durbin-Watson estimate was used to check that there were no significant correlation of the error terms, finally the a Case wise Diagnostic test was used to determine any outliers within an estimation of two and three standard deviations. These statistical tests provided the analysis of the research to give much insight to the MEAP test results of met/exceeded standard in Michigan public elementary, middle, and high school levels for the 2004 and 2005 academic years.
