A Comparison of Social and Tangible Reinforcement with Institutionalized Adolescent Offenders Using Verbal Conditioning Task

The present study is an investigation of the effects of social reinforcement (verbal praise) and tangible reinforcement (tokens redeemable for money) in the learning of a verbal conditioning task by female adolescent delinquents. The study was undertaken in order to determine what environmental events or stimuli serve as effective reinforcers for this population and to assess the effect of presentation method on reinforce effectiveness, as well as to clarify inconsistent findings reported in previous research. Forty Ss, assigned to four groups, were administered a verbal conditioning task using Taffel’s (1955) sentence construction technique. A 21-trail base rate was established for all Ss to make up Block I. Experimental Group I was administered tangible reinforcers during trials 22-63 (Blocks II and III) for sentences beginning with personal pronouns, and social reinforcers presented randomly on pre-determined trials regardless of response. Conditions were reversed for Group II, with social reinforcement given for personal pronouns and tangible rewards presented randomly. To assess the effect of the “masking” procedure, Group III and IV were rewarded contingent upon the correct pronoun response, without presentation of the random reinforcers. Group III received tangible and Group IV social reinforcers for correct responses on trails 22-63 (Blocks II and III). The final 21 trails (Block IV), without reinforcement, served as an extinction block, for all experimental groups. Significant conditioning was demonstrated under all conditions of reinforcement, with an increase in the rate of personal pronoun response form the mean base rate level. Neither type of reinforcement (tangible or social) nor mode of presentation (masked or unmasked) proved to be a significant variable, although a trend indicated that social, unmasked reinforcement might be more effective in changing the behavior of this population. Awareness of the conditioning procedure was not demonstrated, and therefore was deemed to have no significant effect on S performance