Differences between rates of responding emitted during simple and multiple schedules

Abstract

Pigeons pecked keys for food reinforcers delivered by several variable-interval and multiple variable-interval schedules. The rates of responding emitted during the simple schedules were not systematically different from the rates emitted during the multiple schedules when the components of the multiple schedule were identical. The rates of responding emitted during the components were usually greater than the rates emitted during comparable simple schedules when the components were more favorable than the added components of the multiple schedules. Response rates during the components were not significantly lower than those during comparable simple schedules when the components were less favorable. The observation of higher rates of responding during the more favorable components conforms to a prediction of several additive theories (e.g., Rachlin, 1973) but violates a prediction of Herrnstein's (1970) theory. However, the additive theories are brought into question by the fact that changing the location of the discriminative stimuli did not change the pattern of results.Two major theories make different predictions about the relation between the rates of responding emitted during simple and multiple schedules. A theory proposed by Herrnstein (1970) predicts that rate of responding during a simple schedule will be greater than or equal to the rate emitted when the same schedule appears as one component of a multiple schedule. On the other hand, several additive theories (e.g., Gamzu & Schwartz, 1973;Hearst & Jenkins, 1974;Rachlin, 1973) predict that the relation between the two rates of responding may depend on the relation between the rates of reinforcement supplied by the components of the multiple schedule.Herrnstein (1970) proposed that Equation 1 describes the behavior of animals responding on simple, concurrent, and multiple schedules of reinforcement.(1) P I is the rate of responding emitted for a rate of reinforcement equal to R I • R 2 is the rate of reinforcement obtained by making a second response, P 2 , if one is available. The symbols k, m, and R, are parameters that are estimated from the data with certain restrictions (Herrnstein, 1974).

Simple schedule and signal-key multiple schedule responding and behavioral contrast

“…The present experiment examines this question for signal-key multiple schedules. Contrary to the present hypothesis, McSweeney (1980) found that the rates of responding during the more favorable components of signal-key multiple schedules were greater than the rates of responding during comparable simple schedules, even though positive contrast is usually not observed during signal-key multiple schedules. However, McSweeney's (1980) experiment used experienced subjects.…”

“…1974 ;Spealman, 1976. 132 sessions, although 2 of his 6 subjects had previous experience). The one researcher who did find positive contrast and reported the experimental histories of the subjects (McSweeney, 1980, Experiment 3) used subjects who had responded on multiple and VI schedules for more than 1,000 sessions. Therefore, neither differences between simple and signalkey multiple schedule responding nor positive behavioral contrast occurs when subjects have had little experience, but both do develop later.…”

“…, a multiple VI x VI x schedule). McSweeney (1980) asked whether contrast also could be measured relative to a simple schedule baseline; that is, similar to positive contrast, responding might be faster during the more favorable component of a multiple schedule than during the same schedule appearing alone (see, e.g., Freeman, 1971; Halliday & Boakes, 1974;Hearst & Gormley, 1976;Jaffe, 1973;McSweeney, 1980;Pear & Wilke, 1971). Similar to negative contrast, responding might be slower during the less favorable component than during the same simple schedule (e.g., McSweeney, 1982).…”

Herrnstein’s equation for the rates of responding during concurrent schedules

“…Contrary to what Herrnstein's theory predicts, Edmon (1978) did not find an increase in the rates of responding as the components of the multiple schedule got longer. Several studies nave also shown that Herrnstein's theory makes incorrect predictions about the relation between simple and multiple-schedule responding (see McSweeney, 1980). As a result, Herrnstein's theory does not provide an adequate description of responding on concurrent and multiple schedules.…”

Another Look at Contrast in Multiple Schedules

“…In spite of its ability to account for a variety of different findings, Equation 9 also makes several incorrect predictions, as noted by several investigators (deVilliers, 1977;Edmon, 1978;McSweeney, 1980;McSweeney & Dericco, 1976;Spealman & Gollub, 1974 One problem that remains even with Equation 10 is the effect of absolute rate of reinforcement on the magnitude of contrast. As noted by Spealman and Gollub (1974) and McLean and White (in press),Equation 9 predicts that contrast should be greater the higher the rate of reinforcement, and this is true for Equation 10 as well.…”