Alternative accounts are preferable to value transfer theory: Commentary on Dorrance, Kaiser, and Zentall (1998)

Aitken (1999) It has been proposed that, in a simultaneous discrimination, some of the value of the positive stimulus transfers to the negative stimulus (Fersen, Wynne, Delius, & Staddon, 1991). Since then, considerable support for value transfer in pigeons has been reported (Clement, Weaver, Sherburne, & Zentall, 1998; Dorrance, Kaiser, & Zentall, 1998; Dorrance & Zentall, 1999;Zentall & Sherburne, 1994; Zentall, Sherburne, Roper, & Kraemer, 1996).The typical procedure for demonstrating value transfer involves training with two simultaneous discriminations in which the probability ofreinforcement associated with the two positive stimuli is different (e.g., AlOOBO and C50DO, where the letters represent arbitrary stimuli counterbalanced to control for stimulus preference and the numbers following each letter represent the probability ofreinforcement associated with responses to each of the stimuli). Following training, when the pigeons are given a choice between the two negative stimuli (B and D), Preparation of this article was facilitated by Grants MH55ll8 and MH59194 from the National Institute of Mental Health. Send correspondence to T. R. Zentall, Department of Psychology, University of Kentucky, Lexington, KY 40506-0044 (e-mail: zentall@pop.uky.edu). they typically choose the one (B) that has been paired in training with the higher valued positive stimulus (A).To argue against the possibility that the preference for B over D might result from the greater number of nonreinforced responses to D than to B, we have noted two findings: First, perhaps surprisingly, the proportion of total "errors" made to D is generally not significantly different from the proportion of total errors made to B. Acquisition is generally rapid and preexperimental stimulus preferences account for many of the errors made. Second, the correlation between the proportion of errors made to D and the magnitude of the preference from B is typically negative. One might expect it to be positive if differential inhibition to the two negative stimuli were responsible for the preference for B over D that has been found.Aitken (1999) has argued that the evidence against differential inhibition as an account of the effects found is not convincing. First, he notes that other individual differences unrelated to differential inhibition may account for the within-experiment negative correlation between proportion of responses to D in training and the preference for B in test. Second, he notes that, between experiments, there is actually a positive correlation between proportion of responses to D in training and choice of B in test. Third, he proposes that residual evidence for value transfer may result from stimulus generalization between the positive and negative stimuli that is produced by their physical similarity rather than by their simultaneous presentation. We will address each of these criticisms in turn. Differential Inhibition in TrainingFirst, as already noted, the two discriminations were acquired at rates that typically did not diffe...

Section: Stimulus Generalizationcontrasting

confidence: 50%

Section: Differential Inhibition In Trainingmentioning

confidence: 99%

Section: Negative Correlation Between Proportion Of Responses To D Inmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Differential inhibition and stimulus generalization cannot account for value transfer in simultaneous discrimination learning by pigeons: Reply to Aitken

Zentall

Dorrance

Clement

1999

“…Thus, although there is evidence that single-stimulus presentations of the S1 tend to weaken the association between the S1 and the S in a simultaneous discrimination, if one controls for this effect by equating it for the two discriminations, the value that the S1 can transfer to the S depends on a combination of its value when paired with the S , together with its value in other contexts. Aitken (1999) has recently argued that what appears to be value transfer in this and earlier studies actually results from differential inhibition that accrues to the two S stimuli (see also Couvillon & Bitterman, 1992;Wynne, Fersen, & Staddon, 1992). The argument is that 100% reinforcement to A 100 in the A 100 B 0 discrimination should lead to few choices of B, whereas partial reinforcement of responding to C 50 in the C 50 D 0 discrimination should lead to more choices of D. If pigeons make more choices of D 0 than of B 0 in training, more inhibition should develop to D than to B, and on BD test trials, B should be preferred.…”

Section: Value Transfer Versus Occasion Settingmentioning

confidence: 99%

Simultaneous discrimination learning: Stimulus interactions

Zentall

Clement

2001

Stimulus Similarity EffectsThe degree of physical similarity between the S1 and the S during the acquisition of a discrimination has played an important role in theories of discrimination learning (see, e.g., Lawrence, 1955;Logan, 1966;Spence, 1937). By making certain assumptions-(1) that the excitatory strength of the S1 is greater than the inhibitory strength of the S ; (2) that the shape of the generalization gradient is concave downward; and (3) that the net effect can be represented by the algebraic summation of the two (i.e., excitation minus inhibition)-Spence was able to 311Copyright 2001 Psychonomic Society, Inc.Preparation of this article was supported by Grants MH55118 and MH59194 from the National Institute of Mental Health to T.R.Z. We thank Lou M. Sherburne, Brigette R. Dorrance, Janice E. Weaver, Daren H. Kaiser, and Andrea M. Lonon for their contribution to the research described here and William Timberlake for his encouragement to prepare this article. Correspondence concerning this article should be sent to T. R. Zentall, Department of Psychology, University of Kentucky, Lexington, KY 40506 (e-mail: zentall@pop.uky.edu). Simultaneous discrimination learning:Stimulus interactions THOMAS R. ZENTALL and TRICIA S. CLEMENT University of Kentucky, Lexington, KentuckyAlthough considerable attention has been given to interactions between events serving as the positive (S1) and negative (S ) stimuli in successive discriminations, there has been little study of similar interactions in simultaneous discriminations. We propose that in a simultaneous discrimination, given what is typically relatively little experience with the consequences of responding to the S , some of the value of the S1 transfers to the S with which it was paired. Furthermore, the mechanisms responsible for this transfer of value appear to be the higher order Pavlovian association between the S and the S1, as well as within-event associations between them. Although in typical simultaneous discriminations, negative value does not appear to transfer from the S to the S1, when adequate experience is provided with the S , contrast typically develops, reducing the value of the S (negative contrast) and enhancing the value of the S1 (positive contrast). This model of stimulus interactions has implications not only for simple simultaneous discrimination learning, but also for research using combinations of discriminations (e.g., transitive inference in animals and humans). This model may also have implications for a number of human social psychological phenomena.

Value transfer in simultaneous object discriminations by rats

Cohen

Drummond

Terrelonge

2001

) with pigeons trained on two concurrent simultaneous color discriminations (A1B , C1D ) in which the probability of reinforcement for responding to A was greater than that for responding to C. When the birds were tested for their B versus D preferences, they consistently chose B, the stimulus paired with the more positive A, over D, the stimulus paired with the less positive C. This effect was observed whether A was made more positive within the A1B and C1D discrimination context (Zentall & Sherburne, 1994;Zentall et al., 1996, Experiments 1 and 2) or on off-baseline trials (Dorrance et al., 1998; Zentall et al., 1996, Experiment 3). Dorrance et al. noted that this effect is contrary to the occasion-setting hypothesis, which predicts that birds should avoid B (and thus choose D) because the former had come to signal a decrease in reinforcement to its associated A stimulus.Aitken (1999) reanalyzed Dorrance et al.'s (1998) data to argue that the pigeons' preferences for B over D could have resulted from the development of greater conditioned inhibition to D than to B during initial discrimination training. Aitken (1999, see Figure 1, p. 292) found a strong direct relationship (r 2 5 .93) between B preference on BD tests and overall errors to D during discrimination training. It is noteworthy that Dorrance et al. found that their birds consistently made more errors to D than to B during training, except in Experiment 3, where their birds failed to reliably prefer B to D. Although they reported a nonsignificant difference between B and D errors in their second experiment, 6 of their 8 pigeons made more D than B errors, 1 showed the same number of errors, and another made more errors to B than to D. Analysis of these data by a Wilcoxon test revealed significantlymore D than B errors (T 5 5.5, p , .05, one-tailed). Earlier studies from their laboratory (Zentall & Sherburne, 1994;Zentall et al., 1996, Experiment 1) also reported tendencies, albeit nonsignificant ones, for more D than B errors.Notwithstanding Zentall, Dorrance, and Clement's (1999) reply to Aitken (1999), we believe that differential experiences with the negative stimuli during training must Value transfer in simultaneous object discriminations by rats JEROME S. COHEN, CHRISTINE DRUMMOND, and NICOLE TERRELONGE University of Windsor, Windsor, Ontario, Canada Increasing the amount of sunflower seeds under a baited object when it occurs either within or external to a simultaneous object discrimination increases the value transferred to the nonbaited object. Rats were trained with two object discriminations, A 1 B 0 and C 1 D 0 , where the superscript refers to the number of sunflower seeds obtained for pushing aside the object. The amount under A was increased to six on other trials in which the A 6 or C 1 object alternated with regular A 1 B 0 or C 1 D 0 discrimination trials. The A and C objects on the alternating trials were presented alone (Experiment 1), with a common nonbaited object (A 6 X 0 and C 1 X 0 ; Experiments 2 and 3), or with two diffe...