Four types of research on superstitious behavior in humans and nonhumans were reviewed and discussed from the perspective of recent developments in this area. The first type explains the effects of periodic delivery of food on pigeons' superstitious responding in terms of stimulus substitution or species-specific appetitive behavior. Though this type of research is sometimes said to replicate Skinner (1948)'s experiment, this statement may not be accurate because the experiments emphasize the effects of contingencies in steady-state performance. The second type examines whether or not idiosyncratic and stereotyped patterns of behavior develop under response-independent contingencies with attention to accidental response-reinforcer contiguities. The third type includes a variety of studies examining the effects of response independence, such as delay of reinforcement. The fourth type studies superstitious behavior under response-dependent contingencies. It is suggested that further studies should be designed to examine more precisely the effects of response-reinforcer contiguity, aversive control, accidental reinforcement under response-dependent schedules, and verbal control over human superstitious behavior.

View full abstract

This case study was planned to develop the mand and self-initiating behavior in a mentally retarded child who couldn't use demanding speech in her daily life. In order to achieve that purpose, a play equipment, "swing" which seems to have a function of strong reinforcement, was used and the therapist provided mand model and verbal cue, and gradually faded out them by sessions. After 21 therapeutic sessions, their effects upon the child were examined. During baseline, the client chose a swing to play for longer time than any other equipment in the playroom. Therefore it seemed to have a function of the strongest reinforcement. Then the swing was raised up to the point where the client couldn't get on by herself but touched it. When the client approached to the swing, therapist showed her verbal cue "nani ?" which meaned "What do you (client) want to me (therapist) ?" If she didn't make any response, the therapist gave a mand model "yatte" which meaned "Whould you (therapist) kindly help me (client) to get ride on the swing ?" This cue and mand model were faded out gradually by sessions. As a result, the client could request "yatte" to get on the swing without verbal cue and mand model. These findings seem to prove the effectiveness of incidental teaching combined with mand model procedure to shape mand behavior in mentally retarded children.

View full abstract

Dcum is an off-line program for drawing cumulative records on the CRT display. Dcum reads data from disk files and draws cumulative records with such options as changing the scale, compressing the records, synchronizing two records. The source and executable code, written in MS-C for the NEC PC-9801 or compatible computer, are available from the author.

View full abstract