Preference and resistance to change in concurrent variable-interval schedules

Abstract

Response strength is a theoretical construct that has been related to empirical measures in a variety of ways. Skinner (1938) proposed the concept of reflex reserve, which he measured by resistance to extinction.Later, after abandoning reflex reserve (e.g., Skinner, 1950), he invoked probability of respondingas the measure of strength,with response rate serving as a surrogate for probability. Herrnstein (1970) proposed choice behavior as a more appropriate measure, on the basis of his demonstration of orderly relationships between choice behavior and reinforcement rates for concurrent schedules (the matching law), which he then extrapolated to simple and multiple schedules of reinforcement. Nevin (1974) proposed yet a third measure of response strength: resistance to change. Nevin (1974) defined response strength by using the concept of behavioral momentum, which implies that the strength of behavior is indexed by the persistence of responding in the presence of various disrupters (e.g., additional response-independent food). A critical issue for behavior analysis is to identify the relationshipbetween these different measures of response strength. Moreover, to the extent that they do not covary, which measure is more useful for determining the fundamental laws of behavior?Dissociations between response rate and preference and between response rate and resistance to change have been demonstrated in studies of anticipatory behavioral contrast. Williams (1991Williams ( , 1992 presented pigeons a four-component multiple schedule in which two target components(A and B) with identical reinforcement schedules were followed differentially by either extinction or a higher rate of reinforcement. Response rate was higher during ComponentA, which was followed by extinction. However, in probe tests in which the stimuli correlated with Component A and Component B were presented together, preference was in favor of the stimulus correlated with Component B, which had been followed by the richer schedule. Thus, the increase in response rate produced by the contrast procedure was inversely related to the measure of preference. A similar dissociation between response rate and resistance to change was reported by Nevin, Smith, and Roberts (1987), who used a similar four-component schedule and then measured the degree of response persistence during various disruptors. Here, greater resistance to change occurred for the target component followed by the richer schedule.On the basis of this and other evidence, Nevin (1979Nevin ( , 1992Nevin & Grace, 2000) has argued that resistance to change and preference are strongly, and perhaps perfectly, correlated measures of response strength and that both are more orderly indices of the effects of reinforcement training than is response rate. Nevin has further argued that the critical determinant of resistance to change is the rate of reinforcement in the presence of a stimulus, even when that reinforcement may not be contingent on operant responding in the presence of the stimulus. For exampl...

Are preference and resistance to change convergent expressions of stimulus value?

“…All data points were below 0.0, indicating systematically greater preference for the VI 40(Lean) stimulus. Further, preference in the present study approximated preference obtained in previous studies (Belke, ; Bell & Williams, ; Gibbon, ). To be consistent with these previous studies, we express preference differently here, as a proportion of response rates to the VI 40(Lean) stimulus relative to response rates to the VI 40(Rich) stimulus.…”

“…Bell and Williams () suggest that their findings reveal a limitation to the generality of the relation between preference and resistance to disruption. Bell and Williams, however, assessed preference with brief extinction probes, a different procedure from that used by Nevin and colleagues.…”

“…Bell and Williams () explored the generality of the relation between relative resistance to disruption and preference by assessing preference differently than Grace and Nevin (, ). They arranged a two‐component multiple schedule with two concurrently available responses in both components with pigeons.…”

Differential reinforcement and resistance to change of divided-attention performance

“…Resistance to presession feeding and decreased sample duration were similar between trial types, whereas resistance to extinction was greater for the leaner trial type during the 1:9 and 9:1 conditions. Although they are inconsistent with sensitivity of baseline accuracy, these findings are consistent with previous findings in concurrent-choice situations that have shown no clear pattern of resistance to disruption as a function of different reinforcement rates between two alternatives (e.g., Bell & Williams, 2002; McLean & Blampied, 1995; McLean, Campbell-Tie, & Nevin, 1996; Nevin et al, 1990). While they resembled the present findings with presession feeding and extinction, however, these previous studies nevertheless revealed greater overall resistance to disruption of responding on both alternatives in components arranging higher reinforcement rates.…”

Deprivation level and choice in pigeons: A test of within-trial contrast

“…Here, too, results differ depending on whether one or two stimuli are presented simultaneously. When multiple schedules are used (akin to our Experiment 1), greater resistance to change in the richer schedule is observed, but when training involves concurrent schedules (akin to our Experiment 2), there is no differential resistance to change across the two concurrently presented alternatives, despite differences in reinforcement frequency for the two alternatives (Bell & Williams, 2002).…”