Date of Award


Degree Name

Doctor of Philosophy




Previous research suggests that by making classroom contingencies for verbal responding more like those in the natural environment, i.e., by using motivative variables to evoke verbal responses which are followed by specific actions on the part of listeners, verbal repertoires can be taught more effectively to children and language impaired individuals. The present study examines the role of specific versus generalized conditioned reinforcement following adult students' question-asking on the acquisition of a verbal repertoire in a classroom setting and on the generalization of this repertoire to a more natural verbal environment.

Subjects were 32 university students in two American Sign Language (ASL) classes. Following teacher demonstration of vocabulary and grammatical structures, students were provided with a set of questions incorporating those items. In Condition 1, students practiced translating the questions with the teacher, who provided praise for correct responding. In Condition 2, pairs of students practiced asking each other the same questions and writing each other's answers. It was hypothesized that practice which involved asking questions and getting answers (natural reinforcement) would be more effective than practice which involved questions followed by praise (contrived reinforcement), which is typical of more traditional methods of language instruction. All students in one class were exposed to Condition 1 for the first session. Each week thereafter, for 9 weeks, two students w ere randomly assigned to Condition 2. In the second class, the procedure was reversed. Measures of accuracy (number of errors) and fluency (response times) were obtained via weekly tests and an interview with a fluent ASL-user. Nine students received maintenance tests and interviews two months after training.

Test scores revealed little difference in accuracy—Condition 1 subjects averaged 86%, while Condition 2 subjects averaged 89%. Fluency was more strongly affected-subjects required 15% less time to complete tests in Condition 2 than in Condition 1. Arranging test and interview data in order of increasing amounts of exposure to Condition 2 further indicated that the more training students received utilizing the question-and-answer procedure, the more readily their skills generalized to a naturalistic setting and the more vigor with which they displayed their skills in follow-up measures.

Access Setting

Dissertation-Open Access

Included in

Psychology Commons