An analysis of school-based factors that contribute to incarceration rates in Texas counties

The purpose of this study was to identify the relationship between factors found within local education agencies in the 20 most populous counties in Texas and new inmates incarcerated in Texas Department of Criminal Justice (TDCJ) state jail and prison facilities in the same 20 counties. The difference among the dropout rates of three county groups of varying sample sizes was also analyzed.; Data centered upon the factors involved within this study were obtained from three primary data sources: the Texas Education Agency's Academic Excellence Indicator System (AEIS), the Statistical Report for Fiscal Year 2004 published by the TDCJ, and the United States Census Bureau's 2000 census summary in conjunction with 2004 population projections. Significance for the study was set at alpha = .05 for statistical relationships and statistical differences. The major findings of this study were derived from a multiple-regression model, individual correlations between the multiple-regression model's seven individual predictor variables and its sole criterion variable, and a one-way analysis of variance (ANOVA) model. The comprehensive multiple-regression analysis revealed a statistically significant relationship among the seven school-based factors and the percent of new inmates incarcerated in state jail and prison facilities in Texas' 20 largest counties. However, no significant individual relationship was found between the percentage of new inmates in state jail and prison facilities in the 20 counties studied and 6 of 7 of the study's predictor variables.; The variables that were found to have no significant individual correlation included the percentage of students enrolled in bilingual and English as a Second Language (ESL) programs, the average teacher salary, the percentage of the student sample considered at risk, the percentage of the teacher sample with advanced degrees, and the district Texas Education Agency (TEA) rating. The sole predictor variable found to maintain a statistically significant individual relationship with the multiple-regression model's criterion variable was the percentage of students who dropped out from grades 7 through 12. The one-way ANOVA procedure revealed a significant difference between dropout rates in three distinct Texas county groupings based on residential density.