Recent researchers of ABE dropout have commented on the contradictory or inconclusive nature of much dropout research (Darkenwald and Gavin, 1987). Several researchers are attempting to build conceptual frameworks and have conducted theory-based studies of segments of the ABE population. (See for example, Darkenwald and Gavin, 1987; Garrison, 1985, 1987.) Diekhoff and Diekhoff (1984) identified five variables that were related to dropout. Bosma (1988) conducted a study on attrition and persistence with a large sample of adults in a midwestern urban centre over a two year period and confirmed some previous findings.

One of the best known conceptualizations of adult dropout is the "congruence" model developed by Boshier (1973). He suggested that congruence both within the participant and between the participant and his education environment determines dropout/persistence (p. 256). If a learner's inner state of harmony is disturbed (i.e. incongruities develop which result in anxiety), dropout is likely to occur. Boshier measured self/ideal self, self/instructor, and self/other students incongruities and used the scores to predict dropout. In analyzing the variables with persisters and dropouts he found that dropouts showed greater incongruities in self/ideal self and self/others incongruence.

Self/instructor and self/other students are dimensions of the classroom environment. In Boshier's work, however, they were reduced to self/other incongruence and thus conceptualized as a psychological variable. In a critique of the work, Garrison (1987) points out that the model was further reduced to one psychological construct namely, self-esteem. The model has thus been criticized for its psychological reductionism. Its generalizability has also been questioned because it was based on a study using a sample of university continuing education students. Garrison (1985, 1987) has argued for using a more holistic psychosocial perspective in dropout studies and for limiting generalizations to situation-specific populations.

In his 1985 study, Garrison discovered that dropouts thought their courses were more relevant and were more certain of their goals than persisters. This apparently anomalous finding was subjected to further analysis which revealed that "scholastic ability and course relevancy interaction added to the discrimination of persisters and dropouts more than course relevancy itself" (p. 36). Also, a self-confidence and ideal/self discrepancy variable was shown to be a good discriminator of persistence and dropout. These findings were interpreted as showing that ABE learners "with lower scholastic ability, lower self-confidence, and greater socio-economic change may set unrealistic goals for themselves and have unrealistic expectations of the program" (p. 36). Thus, incongruities appear resulting in dropout. These findings have implications for counselling potential learners.