Signal-detection analyses of conditional discrimination and delayed matching-to-sample performance

Quantitative analyses of stimulus control and reinforcer control in conditional discriminations and delayed matching-to-sample procedures often encounter a problem; it is not clear how to analyze data when subjects have not made errors. The present article examines two common methods for overcoming this problem. Monte Carlo simulations of performance demonstrated that both methods introduced systematic deviations into the results, and that there were genuine risks of drawing misleading conclusions concerning behavioral models of signal detection and animal short-term memory.     

Simple and conditional visual discrimination with wheel running as reinforcement in rats.

Three experiments explored whether access to wheel running is sufficient as reinforcement to establish and maintain simple and conditional visual discriminations in nondeprived rats. In Experiment 1, 2 rats learned to press a lit key to produce access to running; responding was virtually absent when the key was dark, but latencies to respond were longer than for customary food and water reinforcers. Increases in the intertrial interval did not improve the discrimination performance. In Experiment 2, 3 rats acquired a go-left/go-right discrimination with a trial-initiating response and reached an accuracy that exceeded 80%; when two keys showed a steady light, pressing the left key produced access to running whereas pressing the right key produced access to running when both keys showed blinking light. Latencies to respond to the lights shortened when the trial-initiation response was introduced and became much shorter than in Experiment 1. In Experiment 3, 1 rat acquired a conditional discrimination task (matching to sample) with steady versus blinking lights at an accuracy exceeding 80%. A trial-initiation response allowed self-paced trials as in Experiment 2. When the rat was exposed to the task for 19 successive 24-hr periods with access to food and water, the discrimination performance settled in a typical circadian pattern and peak accuracy exceeded 90%. When the trial-initiation response was under extinction, without access to running, the circadian activity pattern determined the time of spontaneous recovery. The experiments demonstrate that wheel-running reinforcement can be used to establish and maintain simple and conditional visual discriminations in nondeprived rats.     

On the development and mechanics of delayed matching-to-sample performance

Despite its frequent use to assess effects of environmental and pharmacological variables on short-term memory, little is known about the development of delayed matching-to-sample (DMTS) performance. This study was designed to examine the dimensions and dynamics of DMTS performance development over a long period of exposure to provide a more secure foundation for assessing stability in future research. Six pigeons were exposed to a DMTS task with variable delays for 300 sessions (i.e., 18,000 total trials; 3,600 trials per retention interval). Percent-correct and log-d measures used to quantify the development of conditional stimulus control under the procedure generally and at each of five retention intervals (0, 2, 4, 8 and 16-s) individually revealed that high levels of accuracy developed relatively quickly under the shorter retention intervals, but increases in accuracy under the longer retention intervals sometimes were not observed until 100-150 sessions had passed, with some still increasing at Session 300. Analyses of errors suggested that retention intervals induced biases by shifting control from the sample stimulus to control by position, something that was predicted by observed response biases during initial training. These results suggest that although it may require a great deal of exposure to DMTS prior to obtaining asymptotic steady state, quantification of model parameters may help predict trends when extended exposure is not feasible.     

Emergent, untrained stimulus relations in many-to-one matching-to-sample discriminations in rats

The present experiment investigated whether rats formed emergent, untrained stimulus relations in many-to-one matching-to-sample discriminations. In Phase 1, rats were trained to match two samples (triangle and horizontal stripes) to a common comparison (horizontal stripes) and two additional samples (circle or vertical stripes) to another comparison (vertical stripes). Then, in Phase 2, the rats were trained to match the one sample (triangle) to a new comparison (black) and the other sample (circle) to another comparison (white). In the Phase 3 test, half the rats (consistent group) were given two new tasks in which the sample-correct comparison relation was consistent with any emergent stimulus relations that previously may have been learned. The remaining 6 rats (inconsistent group) were given two new tasks in which the sample-correct comparison relation was not consistent with any previously learned emergent stimulus relations. Rats in the consistent group showed more accurate performance at the start of Phase 3, and faster learning to criterion in this phase, as compared with rats in the inconsistent group. This finding suggests that rats may form emergent, untrained stimulus relations between the discriminative stimuli in many-to-one matching-to-sample discriminations.     

Matching- and nonmatching-to-sample concept learning in rats using olfactory stimuli

Previous research has shown that rats can learn matching-to-sample relations with olfactory stimuli; however, the specific characteristics of this relational control are unclear. In Experiment 1, 6 rats were trained to either match or nonmatch to sample in a modified operant chamber using common household spices as olfactory stimuli. After matching or nonmatching training with 10 exemplars, the contingencies were reversed with five new stimuli such that subjects trained on matching were shifted to nonmatching and vice versa. Following these reversed contingencies, the effects of the original training persisted for many trials with new exemplars. In Experiment 2, 9 rats were trained with matching procedures in an arena that provided for 18 different spatial locations for comparison stimuli. Five subjects were trained with differential reinforcement outcomes and 4 with only one type of reinforcer. Differential outcomes and multiple exemplars facilitated learning, and there was strong evidence for generalization to new stimuli for most rats that acquired several conditional discriminations. Performances with novel samples were generally above chance, but rarely reached the high levels obtained during baseline with well-trained stimulus relations. However, taken together, the data from the two experiments extend previous work, show that rats can learn both match and nonmatch relations with different experimental protocols, and demonstrate generalization to novel sample stimuli.     

Identity matching-to-sample with olfactory stimuli in rats

Identity matching-to-sample has been difficult to demonstrate in rats, but most studies have used visual stimuli. There is evidence that rats can acquire complex forms of olfactory stimulus control, and the present study explored the possibility that identity matching might be facilitated in rats if olfactory stimuli were used. Four rats were trained on an identity match-to-sample procedure with odorants mixed in cups of sand as stimuli. Digging in the sample cup produced two comparison cups, and digging in the comparison cup that contained the same scent as the sample was reinforced. When criterion accuracy levels were reached, novel stimuli were added to the baseline training regimen. All 4 rats reached terminal performance of above 90% correct matching with more than 20 different baseline stimuli and matched novel stimulus combinations with above-chance accuracy; 3 of the 4 rats matched novel stimuli at levels significantly above chance. Accurate matching performance was demonstrated both with 2- and 3-comparison procedures. These results suggest that generalized matching-to-sample can be observed in rats when olfactory stimuli are used and, furthermore, that multiple-exemplar training may be important for its emergence.     

Intertrial sources of stimulus control and delayed matching-to-sample performance in humans

Two experiments compared delayed matching-to-sample (DMTS) accuracy under 2 procedures in adults with mental retardation. In the trial-unique procedure, every trial in a session contained different stimuli. Thus, comparison stimuli that were correct on one trial were never incorrect on other trials in that session (or vice versa). In the 2-sample DMTS procedure, the same 2 comparison stimuli were presented on each trial, and their function changed quasi-randomly across trials conditional upon the sample stimulus. Across 2 experiments, 7 of 8 subjects showed the highest overall accuracy under the trial-unique procedure, and no subject showed consistently higher accuracy under the 2-sample procedure. Negative, exponential decay functions fit to logit p values showed that this difference was due largely to the steeper delay-mediated decline in sample control for the 2-sample procedure. Stimulus-control analyses indicated that, under the 2-sample procedure, the selection of the comparison stimulus on Trial N was often controlled by the comparison stimulus selection on Trial N-1 rather than the Trial-N sample stimulus. This source of competing stimulus control is not present in trial-unique procedures. Experiment 2 manipulated intertrial interval duration. There was a small but consistent increase in accuracy as a function of intertrial interval duration under the 2-sample procedure, but not under the trial-unique procedure.     

The effect of increased response requirements on discriminative performance of the domestic hen in a visual acuity task.

Six domestic hens were trained in a spatial discrimination task. A controlled reinforcement procedure insured that the ratio of scheduled and obtained reinforcement remained equal. Gray stimuli and gratings ranging in spatial frequency from 1 to 10 cycles per millimeter were presented in seven descending series of probes. The response requirement to the sample key was varied from fixed ratio 1 to fixed ratio 40 in seven experimental conditions. An increase in response requirements from fixed ratio 1 to fixed ratio 5 and fixed ratio 10 resulted in significantly higher accuracy at discriminable grating values. Further increases in response requirements did not consistently improve performance. Generally, response biases increased and occasionally became extreme for probes at finer gratings with increased response requirements.     

Quantitative analyses of matching-to-sample performance

Six pigeons performed a simultaneous matching-to-sample (MTS) task involving patterns of dots on a liquid-crystal display. Two samples and two comparisons differed in terms of the density of pixels visible through pecking keys mounted in front of the display. Selections of Comparison 1 after Sample 1, and of Comparison 2 after Sample 2, produced intermittent access to food, and errors always produced a time-out. The disparity between the samples and between the comparisons varied across sets of conditions. The ratio of food deliveries for the two correct responses varied over a wide range within each set of conditions, and one condition arranged extinction for correct responses following Sample 1. The quantitative models proposed by Davison and Tustin (1978), Alsop (1991), and Davison (1991) failed to predict performance in some extreme reinforcer-ratio conditions because comparison choice approached indifference (and strong position biases emerged) when the sample clearly signaled a low (or zero) rate of reinforcement. An alternative conceptualization of the reinforcement contingencies operating in MTS tasks is advanced and was supported by further analyses of the data. This model relates the differential responding between the comparisons following each sample to the differential reinforcement for correct responses following that sample.     

The nature of standard control in children's matching-to-sample

In Experiment I, six preschool-aged children were given matching-to-sample training with two figures in which they were required to choose one of two comparison stimuli that was identical in shape to the standard stimulus. Following this training, they were given intermittent test trials in which a novel stimulus figure was substituted for the previously correct comparison stimulus. Five of the six subjects consistently chose the substituted stimulus during test trials. Experiment II replicated the findings of Experiment I with three other preschool-aged children. Experiment II also provided controls for the possibility that the subjects of Experiment I were selecting the substituted stimulus because of its novelty. The investigators concluded that eight of the nine subjects were exhibiting the type of control described by Berryman, Cumming, Cohen, and Johnson (1965) as S-delta responding.     

Delayed matching-to-sample performance of hens: Effects of sample duration and response requirements during the sample

Six domestic hens were trained under a delayed matching-to-sample procedure with red and green keylights as sample and comparison stimuli and a 1.5-s delay interval. The hens were trained to stop pecking the sample stimuli when a tone sounded. Duration of the sample stimuli (2 to 10 s) and the number of pecks required on the key on which these stimuli were presented (0 to 10) were altered across conditions. Both the response requirement on the sample key and the duration of sample presentations affected accuracy. These findings are in agreement with those of earlier studies using other species and somewhat different procedures.     

Effects of acute and chronic cocaine administration on titrating-delay matching-to-sample performance

The effects of cocaine were examined under a titrating-delay matching-to-sample procedure. In this procedure, the delay between sample stimulus offset and comparison stimuli onset adjusts as a function of the subject's performance. Specifically, matches increase the delay and mismatches decrease the delay. Titrated delay values served as the primary dependent measure. After establishing stable performance in pigeons, several behaviorally-effective doses of cocaine were administered acutely. Dose-related within-session decreases in titrated delay values were observed. Following acute determinations, the dose of cocaine that produced the most rapid decline without eliminating performance was administered prior to each daily session. Chronic administration resulted in performance trending toward control levels. A redetermination of the dose-response function following chronic exposure revealed reduced potency (i.e., tolerance) under cocaine on titrated delay matching-to-sample performance. Supplemental analyses suggest that cocaine may serve as a disruptor of the stimulus conditions in which the performance was established.     

Emergent equivalence relations between interoceptive (drug) and exteroceptive (visual) stimuli.

Conditional "if-then" relations between drug (interoceptive) stimuli and visual (exteroceptive) stimuli were taught to 4 normal humans. Interoceptive stimuli were the effects produced by 0.32 mg/70 kg triazolam (a prototypical benzodiazepine) and placebo (lactose-filled capsules); exteroceptive stimuli were black symbols on white flash cards. Following the training of the prerequisite conditional relations, tests of emergent relations were conducted between exteroceptive stimuli and between interoceptive and exteroceptive stimuli. Equivalence relations emerged immediately without explicit training for all 4 subjects. Accuracy of responding during the interoceptive-exteroceptive equivalence tests and subjects' self-reports showed consistent discrimination between the drug effects of triazolam and placebo. Finally, a generalization test assessed whether a novel visual stimulus presented in the context of the placebo (i.e., no drug) would generalize to visual stimuli belonging to the placebo stimulus class. All 3 subjects who completed this test reliably chose the visual stimuli belonging to the placebo class and not the visual stimuli belonging to the triazolam stimulus class. The development of equivalence relations between interoceptive and exteroceptive stimuli demonstrates that private and public stimulus events can emerge as members of the same equivalence class. Theoretical and clinical implications are discussed.     

Titrating-delay matching-to-sample in the pigeon

The titrating-delay matching-to-sample (TDMTS) procedure offers researchers an additional behavioral task thought to capture some important features of remembering. In this procedure, the delay between sample offset and comparison onset adjusts as a function of the subject's performance. Specifically, correct matches increase the delay and incorrect matches decrease the delay, and steady-state titrated delays serve as the primary dependent measure. The present series of experiments investigated the effects of several procedural variables on performance in TDMTS procedures in an effort to elucidate better its features to allow for more precision in future use. Experiment 1 reports results from a parametric analysis of fixed-ratio response requirements on the sample key that indicated improved remembering in the form of higher daily titrated delay values as the requirement was increased. Experiment 2 investigated the extent to which the initial delay value in each session affected session-wide delay values. Results indicated that regardless of value of the initial delay, the subjects' performances adjusted the delay values in the direction of the known baseline delay-value levels. Experiment 3 manipulated the step size by which delay values were adjusted and the results indicated that larger step sizes increased both session-to-session variability and within-session range of titrated delay values, although the average values remained approximately the same. These results suggest that the TDMTS task serves as a promising procedure to study what many refer to as memory.     

Written lists as mediating stimuli in the matching-to-sample performances of individuals with mental retardation.

Students with mental retardation learned to write lists in order to perform a matching task that they could not do otherwise. After an initial assessment phase, reinforcement was arranged in the computerized tasks to follow selection of the six pictures that were identical to those in the six-picture samples presented. In Study 1, even though the participants wrote a list of the names of the six sample pictures on each trial, read a list, or did both, they often made errors when a brief delay preceded picture selection. In contrast, performance was nearly perfect when a list was written, read, and remained available at the time of picture selection, suggesting that the list served to mediate the delays. Study 2 examined the stimulus control by two- and six-picture samples over the list writing. Early during testing, 1 participant refrained from writing lists on two-picture trials but wrote lists on six-picture trials, thereby maximizing reinforcement and minimizing its delay; the other participant showed this pattern of list writing after supplemental training. The studies suggest methods for establishing a rudimentary repertoire of mediating behavior that has relevance for teaching instruction-following skills in natural settings.     

Seasonal variation in pigeon body weight and delayed matching-to-sample performance

The weights of 5 pigeons with free access to food, monitored over 3 calendar years in the laboratory, were found to fluctuate with season. All pigeons were at their heaviest in the winter and were lightest in the summer. Five different pigeons performed a standard delayed matching-to-sample task for 44 weeks from January to November. Their weights were held at 85% of their summer free-feeding weights, making their predicted deprivation level higher in the winter relative to predicted winter free-feeding weights. Slopes of forgetting functions fit to weekly response totals for each pigeon were shallower in winter, showing an improvement in accuracy with longer delays. Thus, delayed matching-to-sample performance may have been affected by the practice of maintaining the pigeons at a constant body weight throughout the calendar year.     

MATCHING-TO-SAMPLE PERFORMANCE IN RATS: A CASE OF MISTAKEN IDENTITY?

Three rats had previously acquired a simultaneous matching-to-sample performance with steady and blinking lights. In training, the sample stimulus had always appeared on the middle of three horizontally arranged keys with the comparison stimuli on the side keys. In Experiment 1, the sample stimulus appeared on any of the three keys with the comparison stimuli on the remaining two. The matching-to-sample performance broke down with variable sample and comparison locations; the sample stimulus did not control responding to the comparison stimuli when it appeared on a side key, but it retained control when it appeared on the middle key (as in training). In Experiment 2, the rats were trained with the sample always on the left key. When the sample appeared on either of the trained locations (left or middle key), it retained control for both locations. When the sample then appeared on any of the three keys, as in Experiment 1, sample control did not transfer to the untrained location (right key). The experiments demonstrate that training with fixed sample and comparison locations may establish spatial location as an additional controlling aspect of the stimuli displayed on the keys; stimulus location had become part of the definition of the controlling stimuli. The rats' performance seemed best described as specific discriminations involving the visual stimuli and their spatial locations rather than as identity matching.     

Conditional discrimination and equivalence relations: A theoretical analysis of control by negative stimuli

A detailed analysis is presented of the ways in which control by the negative stimulus in two-comparison conditional discriminations may be expected to affect the outcome of tests for the properties of equivalence relations. Control by the negative stimulus should produce the following results: (a) no observable effect on symmetry tests; (b) reflexivity test results should look like “oddity” rather than “identity”; and (c) transitivity tests that involve an odd number of nodes should yield results that are 100% opposite to tests that involve an even number of nodes. The analysis also considers the effects of variation in the type of comparison-stimulus control between and within baseline conditional discriminations. Methods are suggested for experimentally regulating the type of control, and for verifying the predictions that the analysis generates. If suggested experiments continue to support the analysis, investigators who use two-comparison conditional discriminations to study equivalence relations will either have to control explicitly whether the positive or the negative comparison governs their subjects' choices, or they will have to abandon two comparisons and use three or more comparisons instead.     

Delayed stimulus control: recall for single and relational stimuli.

In a discrete-trial symbolic matching-to-sample procedure, pigeons' left-key responses were reinforced following presentation of one center-key sample, and right-key responses were reinforced following presentation of another. Recallability was measured by the difference between log ratios of left to right responses following each sample. In Experiment 1, samples were successively presented same or different wavelengths in the relational discrimination, or individual wavelengths in the single discrimination. The rate at which recallability decreased with increasing delay since sample presentation was the same for single and relational discriminations, but the initial level of performance differed, indicating that the relational discrimination was more difficult. In Experiment 2, recall functions for easy and difficult discriminations between individual wavelengths also differed in levels of initial performance but not in rate of decrement of recallability over time. Recall for stimuli differing in complexity may therefore reflect differences in discrimination difficulty.