Association for Behavior Analysis International

Timeout from positive reinforcement is a procedure in which reinforcement is withheld for a period of time following a target response. Timeout is a common component of behavioral interventions, where it is intended to punish undesired or challenging behavior. The present symposium brings together investigations of timeout from both basic and applied research. The combined research explores experimental parameters, treatment integrity, and release contingencies using a variety of subjects including rats, pigeons, and children. This translational approach to timeout aims to advance the understanding and the application of timeout from positive reinforcement that will ultimately lead to an improved technology of behavior change.

Timeout from positive reinforcement, a common component of behavioral treatments, is used to reduce or eliminate problematic behaviors. We are studying the conditions under which timeout is an effective punisher. Rats lever pressing is maintained on variable-interval schedules arranging reinforcement rates from 0.5 pellets per min to 3 pellets per min. On a conjoint variable-ratio schedule, some presses are followed by a 30-s timeout during which the lever is retracted, a tone sounded, and the food schedule suspended. Under the richest schedule (3 pellets per min) the timeouts do not consistently suppress responding. At leaner schedules, however, timeout does suppress responding. We are continuing to lean the schedules to assess the limits of this effect. These outcomes will help to inform the procedures used in applied settings by clarifying the conditions under which timeout from positive reinforcement functions as an aversive event.

Little is known about how errors when implementing timeout impact timeouts effectiveness. Our experiments address two specific aims: (1) how well do teachers implement timeout with their students, and (2) what are the effects of inconsistent timeout implementation on student behavior. In Experiment 1, we observed teachers ongoing implementation of timeout with students during play. We collected data on the frequency of timeout following targeted problem behavior and non-targeted behavior. Experiment 1 suggested that teachers rarely implement timeout, and when they do, they often do not closely follow their specified procedure. In Experiment 2, we experimentally evaluated the effects of timeout intermittency with the students from Experiment 1. We compared the rates of problem behavior across three integrity conditions: 0%, 100%, and reduced integrity (based on timeout frequency from Experiment 1). So far, Experiment 2 suggests that reduced integrity is about as effective as 100% integrity. These data suggest that teachers may be implementing timeout inconsistently, but as often as needed to suppress target behavior.

Time-out is a commonly used intervention to reduce problematic behavior in young children. Sometimes children engage in problematic behavior during time-out. One suggestion to reduce problematic behavior during time-out is to include a release contingency or require that the child is calm prior to leaving time-out. Previous research has found that adding time to the end of the time-out interval contingent on problematic behavior at the end of time-out is not an effective procedure to reduce problematic behavior during time-out (Donaldson & Vollmer, 2011; Mace et al., 1986). The purpose of the current study is to examine the effects of a release contingency procedure embedded in a 3-min time-out in which each problematic response during time-out results in 15 s of additional time and the absence of problematic behavior for 30 s results in 30 s subtracted from the time-out interval. Thus far, two 4-year-old children, one diagnosed with fetal alcohol syndrome and one with no diagnoses, have participated in the evaluation. Time-out was effective at reducing problematic behavior during enriched time-in contexts (e.g., free play) for both participants. The release contingency reduced or eliminated problematic behavior during time-out for both participants, compared with a fixed duration 3-min time-out.