The present study examined if the opportunity to describe contingent relations after a matching-to-sample conditional-discrimination training procedure differentially affected participants' performances on a generalization task; and if verbal feedback differentially affected the number of trials-to-criterion during training. Eighty college students were randomly assigned to one of four groups (N=20/group). Group 1 received verbal feedback congruent with computer generated contingencies. Group 2 received verbal feedback incongruent with these contingencies. Group 3 received no verbal feedback but was asked to write a description of their correct responses before the generalization test. Group 4 received neither verbal feedback nor the opportunity to write any descriptions. A difference was found between Groups 3 and 4 in the mean number of correct responses in the generalization test. Results suggest that describing experienced contingent relations facilitates the application of rule-following behaviors on subsequent generalization tasks.
