Prediction of adult learner dropout using a psychosocial system model
AbstractStudies of dropout have repeatedly shown a variety of non-school related reasons for not persisting in the formal learning setting. However, some researchers have questioned the validity of those findings. In order to better understand and predict dropout in adult basic education, this study identified two questions. The first concerned whether a hypothesized typological structure of a complex array of psychosocial correlates of dropout was isomorphic with a set of empirical referents. The second question concerned the ability of the independent variables (factor scores) identified in the first part of the study to predict persistence/dropout.
General system theory formed the basis of an abstract social system model to view the adult learner and the corresponding environment. This model in turn was used to help identify the adult learner-environment typology hypothesized in the first stage of the study. The five factors identified were labeled internal and external tensions, capabilities, and internal and external constraints. Variables selected to represent the five categories were the Social Readjustment Rating Scale, Differential Aptitude Test (Verbal), Adjective Check List, Tennessee Self Concept Scale, and two instruments (study time and financial concern) designed specifically for this study. Variables representing the structural categories of the typology were then combined linearly as the resilience model used to predict persistence/dropout. Common factor analysis of the 18 psychosocial variables included in the model led to the identification of five factors congruent with the hypothesized adult learner-environment typology. However, the hypothesized loading of the variables on these five factors showed considerable incongruence with the projected loadings. The variables were therefore reinterpreted in light of the hypothesized typology and the factor analytic results. Calculated factor scores were then used as measures to predict dropout for a sample of adult basic education students. The external tension and constraint variables were significant predictors of dropout accounting for 11.4 percent of the dropout variance. When manifest variables were used as predictors, the Social Readjustment Rating Scale (life changes) and "self control" were found to be significant predictors of dropout and they accounted for 15.6 percent of the dropout variance.
The conclusions were that the adult learner-environment typology showed promise in ordering the many variables and the abstract social systems model demonstrated its usefulness as a philosophical perspective and analytical method. In addition, confirmation was provided for previous research that identified socioeconomic variables as significant barriers to continuation in a formal adult basic education program. The recommendation for the practitioner was to begin to direct attention to variables exogenous to the learning setting. For the researcher, the suggestion was to design studies with both learning setting and psychosocial variables as predictors of dropout in the quest to determine relative predictability as well as attempt to account for a greater percentage of dropout variance than previous research has done.
Education, Faculty of
Educational Studies (EDST), Department of