On the Reliability of Blocking Effects in Response Acquisition with Delayed Reinforcement

Abstract

Background Stimuli perfectly correlated with impending access to food have been shown to both attenuate response rates and to prevent response acquisition when they occur during delays to reinforcement. One explanation of these findings is that the stimulus better predicts food than the operant response itself, and therefore, "blocks" learning of the response-reinforcer association. That such stimuli can abolish operant learning implies a breakdown in an organism's ability to detect causality between its own behavior and effects on the environment. Method Two response acquisition experiments in which a stimulus preceded food delivery were conducted. In one experiment, an attempt was made to replicate the prevention of response acquisition using a non-resetting delay procedure that parallels those that result in overshadowing. In a second experiment, stimulus-food pretraining was given to provide a better parallel to typical respondent-conditioning blocking procedures. Results Under neither circumstance was response acquisition prevented. Discussion The generality and robustness of blocking the response-reinforcer association in operant response acquisition is questioned.Keywords Blocking . Delay of reinforcement . Lever press . Rats . Response acquisition Delay of reinforcement is a primary method by which researchers have sought to answer questions about how the temporal relations among behavior and environmental events combine to affect behavior. In the experimental analysis of behavior, it is well-established that the longer the time between operant response and reinforcer, the weaker the effect of the reinforcer, other things being equal. This relationship becomes less clear when response-reinforcer delays are signaled. For instance, sometimes a delay that nearly abolishes operant behavior when unsignaled has little effect when signaled (Richards 1981). The picture becomes even less clear when one considers that the opposite finding sometimes holds. For instance, Pearce and Hall (1978) found that inserting a 200 ms localized light stimulus at the beginning of a 500 ms response-reinforcer delay reduced lever pressing by Lister rats on a variable-interval (VI) 60 s schedule of reinforcement when compared to a control condition in which the light stimulus was presented randomly. Signals occurring during response-reinforcer delays function differently depending on other parameters of the situation.An example of response attenuation by a delay signal comes from Williams (1975). In this experiment pigeons were first trained with immediate reinforcement to peck a right-side key illuminated green. Next, the pigeons experienced a discrete-trials procedure in which a key-peck during a 5-s left-red keylight presentation resulted in access to grain 9 s after keylight offset with a probability of 1.0. If no peck to the left-red key occurred, grain was delivered with a probability of 0.33 at the end of 9 s. For half of the pigeons, the right-green key was presented during the final 4 s of the delay although pecking this ke...