Thomas DeMarse

On the relationship between differential outcomes and differential sample responding in matching-to-sample.

Four experiments examined control over choice by differential sample responding in matching-to-sample with differential outcomes. In Experiment 1, pigeons initially learned to match with food versus no-food outcomes. Their performances later transferred to other samples to which responding versus not responding had been explicitly reinforced with a single outcome (food). In Experiment 2, pigeons initially learned to produce the comparisons by pecking one sample but not the other. Transfer was then observed to new samples associated with food versus no food (and thus often vs. seldomly pecked). Experiments 3 and 4 showed that transfer of matching required differential behavior to each sample set and did not depend on explicit conditioning of that behavior prior to acquisition. Together, these results show that differential sample behaviour provides a redundant cue for choice in differential outcome matching-to-sample.

Enhancement of matching acquisition by differential comparison-outcome associations

Two experiments investigated whether differential outcomes in matching-to-sample (MTS) would enhance acquisition even in cases in which the outcomes could not be anticipated at the beginning of a trial. In Experiment 1, food and no-food outcomes were differential with respect to both hue and line comparisons in one-to-many matching-to-sample (MTS) but were nondifferential with respect to the samples. Overall acquisition and acquisition with each comparison set were faster in relation to controls that received each outcome equally often on all trials. In Experiment 2, hue and line comparisons were associated with either different probabilities of food (p = 1.0 vs. 0.2) or with the same probability (p =.6). Again, matching acquisition was more rapid in the differential group

Further Tests of Response-Outcome Associations in Differential- Outcome Matching-to-Sample

Two transfer-of-control experiments assessed pigeons sensitivity to response-outc ome associations in differential-outcome discriminations. In Experiment 1, pigeons were trained on one-to-many matching-to-sample with food and no-food outcomes that were differential or nondifferential with respect to correct choice. The samples were then replaced by novel stimuli that had differential or nondifferenti al associations with those same outcomes. Transfer of matching occurred only when the novel samples and their respective choice responses had identical differential-outcome associations. Experiment 2 showed that the outcomes themselves were effective samples if the choices they cued yielded those outcomes in training. These data provide further evidence that the relation between comparison choice and consequent outcomes influences pigeons matching performances. This study follows up our recent provocative finding (Urcuioli & DeMarse, 1996, Experiment 1) that pigeons will readily transfer their differential-outcome conditional discriminations to novel samples even though the samples in training are not associated with different outcomes. In the condition of interest, pigeons learned to match each of two sample stimuli to two different comparison alternatives, with correct choices followed either by food or by a lit food hopper only (no food). These two outcomes were arranged to occur equally often after each sample, although each correct comparison yielded only one outcome. In other words, the comparison alternatives were differentially associated with food and no food in training, but the sample stimuli were not. After acquisition of this task, the birds were tested for their ability to match those comparisons to novel samples that had been differentially associated with food and no food off baseline. The correct comparison on each test trial was designated as the one having the same outcome association as the novel sample preceding it. Under these conditions, we found that birds chose the correct comparison on 80% of the test trials. Table 1 diagrams the training and test conditions that yielded this strong transfer effect. Note, in particular, that each correct comparison (Cn+) in training was followed by only one outcome, whereas the samples (S1 and S2) were followed (on different trials) by both outcomes. The high level of accuracy we observed in testing was unexpected and surprising, given the typical theoretical

Transfer of performance to new comparison choices following differential outcome matching-to-sample

Four experiments examined transfer of differential outcome performances to new choice responses in pigeons. Experiments 1A and 1B showed that new responses trained off a matching-to-sample baseline readily substituted for the choice alternatives in differential outcome matching, provided that they shared the same outcome associations as the alternatives they replaced. Experiment 2 showed that comparison responses trained on baseline, but in a task in which their different outcomes occurred equally often following each sample (viz., one-to-many matching), substituted for the choices in a standard, differential outcome task. Experiment 3 showed, somewhat surprisingly, that the choices in the latter task were likewise effective substitutes in one-to-many matching. These results pose separate challenges for standard two-process theory and for the bidirectional account of differential outcome performance, and they suggest other cues that pigeons may use to predict outcomes.

Assessing the contributions of S-O and R-O associations to differential-outcome matching through outcome reversals.

Pigeons were trained on symbolic matching with 2 samples, 2 pairs of comparisons, and different outcomes for the correct responses within each comparison pair. For one group, the 2 samples were also associated with different outcomes, whereas for another group, they were not. When the response-outcome (R-O) relations for one pair were subsequently reversed, the group trained with differential sample-outcome (S-O) associations was significantly disrupted in its performance on both reversed- and nonreversed-outcome trials. By contrast, the group trained with just differential R-O associations was disrupted only on reversed-outcome trials. These results were replicated when the outcomes on the initially nonreversed trials were then reversed. The findings indicate that differential S-O associations, when present, have a stronger influence on matching performances than differential R-O associations. They are also consistent with hierarchical and configural models of discriminative control.

Control of matching by differential outcome expectancies in the absence of differential sample-outcome associations: a serial compound view.

In this study, pigeons learned 2 separate one-to-many conditional discriminations in which they matched form samples to line and hue comparisons. Correct choices within each comparison dimension yielded differential (food vs. no-food) outcomes that were not predictable from the samples alone. At asymptote, latency to make a correct choice was shorter when food was the contingent outcome than when no food was the outcome. More important, when the samples from each task were subsequently exchanged, comparison choice varied systematically as a function of the sample and the set of new comparison alternatives that followed them. Together, these results indicate that choices were cued by differential outcome expectancies arising from serial compounds consisting of each sample and the dimensional characteristics of the comparisons.

Transfer across delayed discriminations: II. Differences in the substitutability of initial versus test stimuli.

In 2 experiments, pigeons were trained on, and then transferred to, delayed simple discriminations in which the initial stimuli signalled reinforcement versus extinction following a retention interval. Experiment 1 showed that discriminative responding on the retention test transferred to novel test stimuli that had appeared in another delayed simple discrimination but not to stimuli having the same reinforcement history off-baseline. By contrast, Experiment 2 showed that performances transferred to novel initial stimuli whether they had been trained on-baseline or off-baseline. These results suggest that the test stimuli in delayed simple discriminations acquire control over responding only in the memory task itself. On the other hand, control by the initial stimuli, if coded as outcome expectancies, does not require such task-specific training.