Abstract
One of the best known conceptualizations of dropout in adult education is Boshier's congruence model. This model, however, based largely on self concept incongruency variables, has been criticized for its psychological reductionism. In addition, the model's generalizability has been questioned because it was based upon a single study using a sample of university continuing education students. To test the validity of these criticisms and determine the relative power of the incongruency variables in predicting dropout behavior, this study analyzed Boshier's constructs in concert with a variety of other socioeconomic and psychological variables using a population of adult high school completion students. The results revealed that not only were the incongruency variables poor predictors of dropout for this sample of adult learners, but that dropouts had less self/other incongruence. This finding brought into question the adequacy and generalizability of the congruence model in explaining adult dropout behavior. It was suggested that dropout studies be approached from a holistic psychosocial perspective and that generalization be limited to situationally specific populations.