A demonstration of persistent human avoidance in extinction

Contingency knowledge is necessary for learned motivated behaviour in humans: relevance for addictive behaviour

AIMS

Many forms of human conditioned behaviour depend upon explicit knowledge of the predictive contingency between stimuli, responses and the reinforcer. However, it remains uncertain whether the conditioning of three key behaviours in drug addiction-selective attention, instrumental drug-seeking behaviour and emotional state--are dependent upon contingency knowledge. To test this possibility, we employed an avoidance procedure to generate rapidly these three forms of conditioned behaviour without incurring the methodological problems of drug conditioning.

DESIGN

In two experiments, participants (16 students) were trained on a schedule in which one stimulus (S +) predicted the occurrence of a startling noise, which could be cancelled by performing an instrumental avoidance response.

MEASUREMENTS

The allocation of attention to the S + and the rate and probability of the avoidance response in the presence of S + were measured. Following training, participants were tested for their knowledge of the stimulus-noise contingencies arranged in the study and rated the emotional qualities of the stimuli.

FINDINGS

Both experiments showed that S + gained control of selective attention, instrumental avoidance behaviour and subjective anxiety, but only in participants who reported explicit knowledge of the Pavlovian contingency between the S + and the startling noise.

CONCLUSIONS

The implication of the present findings is that the control of selective attention, instrumental drug-seeking behaviour and emotional state by drug-paired stimuli is mediated by cognitive knowledge of the predictive contingency between the stimulus and the drug.

Acquisition, extinction and temporal discrimination in human conditioned avoidance

Two experiments were conducted with the aim of designing a videogame for the study of human conditioned avoidance. Participants had to destroy enemy spaceships with the goal of increasing the score in a counter. Coloured signals might announce the launching of a bomb that could hit participant's spaceship producing a 30 points decrease in participant's score. Three groups of participants were trained in discriminating between a warning signal (W) and a safety signal (S) in Experiment 1. Instrumental group could avoid the loss of points by hiding the spaceship before the offset of W. Participants in the Yoked group received the same treatment received by their instrumental partners, regardless of their behaviour. In the Pavlovian group, W was always followed by the loss of points, regardless of participant's behaviour. Discrimination between W and S was better in the Instrumental groups than in the Yoked and Pavlovian control groups. Experiment 2 found extinction of avoidance when the warning signal was not followed by the bomb. Temporal discrimination was found within the participants that received the instrumental contingency in both experiments, with higher avoidance response towards the end of the warning signal. Temporal discrimination disappeared after extinction in Experiment 2.

The S-S construct of expectancy versus the S-R construct of fear: which motivates the acquisition of avoidance behavior?

The authors provided a differential test between stimulus-stimulus (S-S) and stimulus-response (S-R) theory predictions in regard to the roles that the constructs of expectancy and of fear play in maintaining classically conditioned fear responding within the context of a human conditioned-avoidance paradigm. After the participants had developed sustained avoidance responding, their shock electrodes and avoidance response apparatus were removed to enhance the cognitive expectancy that the conditioned stimulus (CS) would not be followed by the unconditioned stimulus (UCS). This manipulation of expectancy was successful in 96% of the participants. The study was conducted over a 2-day period and involved 1 experimental group and 3 control groups. During the test trials, the authors used autonomic and self-report indices of fear to assess the presence or absence of fear to the CS. The data disconfirmed the prediction of the S-S theory that fear to the CS would be extinguished. The authors discuss the implications of this finding for S-S theories and for approaches in cognitive behavior therapy.