Predicting Drop-Outs in Adult Education Courses

Abstract

The subjects in the study were 104 male adults who were partici pating in an adult retraining program. A drop-out rate of 55.2 per cent was found. The subjects were given tests of intelligence, reading ability, and personality, and certain personal data was also obtained. The drop-outs were less intelligent and had lower read ing ability but did not differ in personality from the completions. The variables having the highest correlation with success were all measures obtained on the personal data form, a total of thirteen of these variables being significantly correlated with success. These variables formed a "nonsuccess" syndrome. The technique of multiple linear regression was extremely useful in analyzing the data, enabling 86.5 percent of the trainees to be correctly pre dicted as drop-outs or completions.