An analysis of timeout and response cost in a programmed environment.

A group of mildly retarded adolescents with high rates of antisocial behavior was exposed to two parameters of timeout and response cost within the context of a programmed environment. For five of the six subjects, the two higher values (30 tokens response cost or 30 min timeout) were significantly more suppressive than the lower values (five tokens or 5 min). For the one remaining subject, there was a strong relationship in the opposite direction. Also, the timeout and response cost of higher value became increasingly more suppressive over time, whereas those of lower value did not. There were few appreciable differences between the timeout and response cost of similar magnitude. A discussion of these results is presented in support of the notion that the functional aversiveness of timeouts (and response costs) appears to be critically dependent upon interactions with the environmental conditions in which they are implemented and the reinforcement histories of the subjects.