The purpose of this study was to examine personal, family, and curriculum variables among high school dropouts with mild disabilities. The review of research literature on dropouts from general and special education assisted in the selection of the variables. The outcome variable (high school dropouts with mild disabilities) and the relationship to the following variables: personal (age, attendance, gender, and ethnicity); family (parents' economic level, parents' educational level, and one vs. two-parent households) and curriculum (academic credits, vocational credits, support credits, and work experience credits) were examined. The research examined the relationship of selected variables to dropouts. Dropout status among students with mild disabilities was found to be 29 percent. When age was examined, increased age was weakly associated with a higher likelihood of becoming a dropout. Increased absenteeism was moderately associated with a higher likelihood of dropping out. However, the strength of the relationship between gender and the drop out status was not significant across any of the three categories of mild disabilities. The analysis of ethnicity found that African-Americans with learning disabilities were more likely to drop out. In addition, the drop out rates for Caucasian youth were statistically significantly lower than drop out rates for other groups. Higher economic level was associated with a lower likelihood of dropping out. Increased educational level and households with two parents were also associated with a lower likelihood of dropping out among students with learning disabilities or mental retardation. The higher the number of credits, the lower the likelihood of dropping out. Dropouts took fewer credits which is not surprising since dropouts by nature have fewer credits. / Ph. D.

Using factors of low income, parents' levels of education, and family composition as determinants of educationally at-risk status, study investigated differences between first generation, undergraduate college students from families in lowest quintile of income in the U.S, One group consisted of students from single-parent households and the other of students from two-parent households. Data were from CIRP 2003 College Student Survey (CSS) and its matched data from the Freshman Survey (Student Information Form - SIF). Differences examined included student inputs, involvements, outcomes, and collegiate environments. Included is portrait of low income, first generation college students who successfully navigated U.S. higher education. The number of cases dropped from 15,601 matched SIF/CSS cases to 308 cases of low income, first generation college students (175 from single-parent households and 133 from two-parent households). Most of the 308 attended private, 4-year colleges. Data yielded more similarities than differences between groups. Statistically significant differences (p < .05) existed in 9 of 100 variables including race/ ethnicity, whether or not English was first language, and concern for ability to finance education as freshman. Data were not generalizable to all low income, first generation college students because of lack of public, 4-year and 2-year colleges and universities in dataset. Graduating seniors' average expected debt in June 2003 was $23,824 for students from single-parent households and $19,867 for those from two-parent households. 32% from single-parent households and 22% from two-parent households expected more than $25,000 of debt. Variables used on SIF proved effective tools to develop derived variables to identify low income, first generation college students from single-parent and two-parent households within CIRP database. Methodology to develop derived variables is explained.