| The purpose of this study was to investigate the complex interrelationships between the psychological constructs of identity status, locus of control, and ego development and student dropout decisions as well as the proposed mediating effects caused by the environmentally oriented variables of GPA and levels of satisfaction in college.;Significant relationships were found to exist between: (1) the psychological construct of identity status and each of the additional psychological variables; (2) the mediating variables of GPA and levels of social and academic satisfaction in college; and (3) certain one-to-one pairings of psychological and mediating variables.;While direct relationships were not established between the psychological variables and student dropout decisions, significant associations do exist between the mediating variables and decisions to persist or withdraw.;These patterns of associations between variables suggest that while the psychological constructs are not directly related to student dropout decisions, they may have been of critical indirect influence. More specifically, individuals at varying stages of development are likely to differ in their processing of and openness to such stimuli. As a result, it would follow that individuals at different developmental levels will differ in the meaning, structure, and direction they attach to these stimuli, accounting for individual differences in perception and environmental impact. Thus, in serving to mediate and possibly control to a considerable degree the effects of the environment on the individual (Reid, 1974), these developmental, psychological variables may have helped to establish a link between the person, the college environment, and the drop-out process.;Data were generated through the administration of a Student Information Form containing (1) a self-designed questionnaire; (2) Loevinger's Washington University Sentence Completion Test; (3) Adams' Objective Measure of Ego-Identity Status; and (4) an 18-item version of Reid and Ware's Multi-dimensional Internal-External Scale. The sample population, drawn from the Massachusetts State Colleges, was comprised of students in April of their freshman year. Usable responses were obtained from 339 students. Descriptive, contingency, and path analyses were applied to the data to test the hypotheses which had been advanced in both null and alternative forms. |
