Derived insensitivity: Rule-based insensitivity to contingencies propagates through equivalence

Effects of reinforcement and response-cost history on instructional control

The present study compared the effects of reinforcement or punishment versus no additional consequences for instruction following on instructional control and subsequent rule-governed insensitivity. In two experiments, adult participants were presented with repeated choices between a short progressive-time schedule and either a fixed time-schedule or a longer progressive-time schedule. In Experiment 1, three groups were given an initially accurate instruction relative to the direct contingency. A control group experienced no additional consequences for compliance with instructions, whereas compliance resulted in additional points for a second group, and noncompliance led to the subtraction of points for a third group. In a subsequent phase, instructions became inaccurate and there were no additional consequences for compliance or noncompliance for any group. Consistent with previous results, rule-governed insensitivity was observed in all participants. Experiment 2 employed the same procedure, except instructions were inaccurate throughout all sessions, and compliance in the subsequent phase resulted in diminishing points per session. Reinforcement for following instructions increased instructional control and subsequent rule-governed insensitivity. This increase was maintained even after the termination of additional consequences, a result that supports theoretical suggestions that a history of reinforcement for complying with instructions and rules is an important factor in rule-governed insensitivity.

The rule-based insensitivity effect: a systematic review

Background

Adherence to inaccurate rules has been viewed as a characteristic of human rule-following (i.e., the rule-based insensitivity effect; RBIE) and has been thought to be exacerbated in individuals suffering from clinical conditions. This review intended to systematically examine these claims in adult populations.

Methodology

We screened 1464 records which resulted in 21 studies that were deemed eligible for inclusion. Each of these studies was examined to determine: (1) if there is evidence for the RBIE in adults and (2) if this effect is larger in those suffering from psychological problems compared to their non-suffering counterparts. In addition, we investigated how (3) different operationalizations of the RBIE, and (4) the external validity and risks of bias of the experimental work investigating this effect, might influence the conclusions that can be drawn from the current systematic review.

Results

(1) Out of the 20 studies that were relevant for examining if evidence exists for the RBIE in adults, only 11 were eligible for vote counting. Results showed that after the contingency change, the rule groups were more inclined to demonstrate behavior that was reinforced before the change, compared to their non-instructed counterparts. Critically, however, none of these studies examined if their no-instructions group was an adequate comparison group. As a result, this made it difficult to determine whether the effects that were observed in the rule groups could be attributed to the rules or instructions that were manipulated in those experiments. (2) The single study that was relevant for examining if adults suffering from psychological problems demonstrated larger levels of the RBIE, compared to their non-clinical counterparts, was not eligible for vote counting. As a result, no conclusions could be drawn about the extent to which psychological problems moderated the RBIE in that study. (3) Similar procedures and tasks have been used to examine the RBIE, but their precise parameters differ across studies; and (4) most studies report insufficient information to evaluate all relevant aspects affecting their external validity and risks of bias.

Conclusions

Despite the widespread appeal that the RBIE has enjoyed, this systematic review indicates that, at present, only preliminary evidence exists for the idea that adults demonstrate the RBIE and no evidence is available to assume that psychological problems exacerbate the RBIE in adults.

The systematic review was registered in PROSPERO (CRD42018088210).