Two-choice Discrimination Learning Research Articles

A new protocol for investigating visual two-choice discrimination learning in lizards

One of the most widely studied abilities in lizards is discrimination learning. The protocols used to test lizards are often novel or adapted from other taxa without proper validation. We need to ensure that tests of discrimination learning are appropriate and properly applied in lizards so that robust inferences can be made about cognitive ability. Here, we present a new protocol for testing lizard discrimination learning that incorporates a target training procedure, uses many daily trials for efficiency and reinforcement, and has a robust, validated, learning criterion. We trained lizards to touch a cue card using operant conditioning and tested lizards separately on a colour, and pattern discrimination test. Lizards successfully learnt to touch a cue card and to discriminate between light and dark blue but had issues discriminating the patterns. After modifying the test procedure, some lizards reached criterion, revealing possible issues with stimulus processing and interference of generalisation. Here, we describe a protocol for operant conditioning and two-choice discrimination learning in lizards with a robust learning criterion that can help researcher better design future studies on discrimination learning in lizards.

Does Joystick Training Facilitate Relational Learning?

Thirteen naïve capuchin monkeys (Cebus [Sapajus] apella) were manually tested with the Transfer Index procedure, a species-fair paradigm for assessing the capacity to learn and to transfer learning. The animals were then trained to manipulate a joystick to control a cursor and to respond to stimuli on a computer screen. After the animals had mastered the remote cause-effect relations required by the computerized test system, they were returned to manual Transfer Index testing to determine whether the joystick-training intervention had affected the monkeys’ capacity for efficient and relational learning. Transfer Index scores and overall accuracy was higher following the joystick intervention, but these differences were not statistically significant. Two-choice discrimination learning and reversal appeared to be associative in nature, and there was no evidence that joystick training made the monkeys more rule-like or relational in their learning. Despite the absence of significant differences, the patterns of results encourage further study of the ways that changes in the cognitive competencies of nonhuman animals might be catalyzed by significant learning experiences.

Effects of methylphenidate on attention in children with moderate mental retardation

Six children with moderate mental retardation participated in a double-blind, placebo-controlled, crossover study of the effects of methylphenidate on performance on simultaneous two-choice discrimination learning problems, a sustained attention task, and classroom activity. Dependent measures were compared for: (1) % correct on discrimination learning problems; (2) sensitivity and bias on a sustained attention task; and (3) time on-task during classroom activities. Only the sustained attention task and the classroom measure showed a clear effect of methylphenidate. Two children performed better on the drug, one performed better on the placebo, and three showed no systematic effects. The two children who responded favorably to the drug spent more time on task in the classroom, showed higher sensitivity on the sustained attention task and adopted a more efficient, less conservative, criterion on this test. These results are inconsistent with the idea that methylphenidate improves the manageability of some children by decreasing behavior levels, in general, or by reducing impulsivity, in particular. Methylphenidate increased response output and the degree of environmental control of this output. The improvement in sensitivity demonstrates an attentional effect that cannot, in this case, be confounded by response bias factors.

Learning and transfer of identity-difference relationships by infants

Seven-month-old infants were presented three successive two-choice discrimination learning problems in which the stimuli were pairs of toys. On each trial, one pair of two identical toys and one pair of two different toys were presented. The infants were reinforced for fixating the positive pair by hearing a tape-recorded story. The relationship (either identity or difference) among the members of the positive pair was constant for the first two problems, and the percentage of total fixation time directed to the positive pairs increased over trials. Infants for whom the positive intrapair relationship was reversed for Problem 3 experienced an immediate decrease in their level of performance, whereas no such decrement occurred for the others. These results were interpreted as indicating that the infants had learned to attend and respond to intrapair relationships.

The informativeness and frequency of presentation of a social stimulus as determinants of its efficacy as a reinforcer

It was demonstrated that variations in the informativeness of a social stimulus place important limitations on the generality of the proposition that deprivation heightens, while satiation lowers, responsiveness to a social stimulus. Using 60 children 5 to 6 years of age as subjects, a social stimulus (“good”) was presented at either a high or a low frequency in one of three information-value contexts (high, medium, or low). Immediately following the exposure treatment the same social stimulus was employed as a reinforcer of nonpreferred responses in a two-choice discrimination learning task. The findings indicated that the heightened efficacy of deprivation relative to satiation only held for low information value; when initial information value was high the subsequent efficacy of a social stimulus was enhanced by its initial recurrent usage. Trend analyses revealed that the effects of variations in information value were strongest within the high-frequency-of-praise treatment. Possible mechanisms underlying the significant Information Value × Frequency interaction were discussed.

Effect of cue salience on discrimination learning

Salience for specific dimensional cues was assessed in second-grade children immediately prior to a two-choice discrimination learning task. The cue reinforced in the learning task was either the salient cue (learned fastest), the alternate cue of the salient cue’s dimension, or a cue from a dimension not salient (learned slowest). Children without specifically salient cues learned in approximately as many trials as did the children who were reinforced for choices of the salient cue or dimension. The results indicated that specific cue salience did not have a general facilitative effect, but that it was related to discrimination learning in a logical and predictable manner.

The effects of stimulus movement on discrimination learning by rhesus monkeys

Two experiments were conducted with test-sophisticated adult rhesus monkeys to determine effects of differentiating and nondifferentiating movement cues on two-choice discrimination learning. In Experiment 1, 12 monkeys were given initial learning problems followed by intradimensional shift problems with movement- and color-differentiated stimulus cues. Results indicated that original learning progressed most rapidly with movement cues, but that subsequent task-shift performance was comparable for movement- and color-cue problems. There was no evidence that learning with either cue dimension transferred to subsequent problems involving the other cue dimension. In Experiment 2, 14 monkeys were given a series of two-choice discrimination problems in which discriminanda differed in color only. Problems were presented under two conditions: (1) both discriminanda moved synchronously during the choice period or (2) both discriminanda remained stationary. Results indicated that the monkeys learned more rapidly when discriminanda were moving rather than stationary. It was concluded that motion is a dimension of comparable saliency to color, although the psychophysical properties need further clarification. It was hypothesized that stimulus movement enhances discrimination learning through facilitation of the appropriate allocation of attention.

The effects of a marijuana extract on two-choice discrimination learning in the squirrel monkey

Eight adult male squirrel monkeys were given acquisition training on two-choice visual discrimination problems on form and color dimensions. Discrimination accuracy on both dimensions was impaired by.68 mg/kg of delta-9-tetrahydrocannabinol pretreatment. Response time was impaired on the first discrimination problem but evidence of tolerance development was present on the second. The accuracy impairment failed to show any evidence of tolerance development.

Wechsler Picture Arrangement Scores and Children's Problem Solving

Summary Sixty-four fifth-grade boys and girls were administered a two-choice discrimination learning task. The two major factors were (a) high versus low picture arrangement scores and (b) two treatment conditions: neutral and cue. Contrary to prediction, children scoring high in picture arrangement demonstrated significantly poorer discrimination learning ability across conditions than children with low picture arrangement scores. The need for further research on the factors underlying picture arrangement performance was stressed.

Arousal and Learning, and Noncontingency Detection in One-Year-Old Infants

The relationship between arousal and learning was studied with 6 infants 1 yr. of age on a two-choice discrimination learning task. The behavioral indices of arousal (activity level) and learning (response rate and reinforcements) correlated significantly (r, .44). Affective state correlated positively with activity level (r, .47) during the learning phase of the experiment. Two control conditions were presented following the learning phase of the experiment. Noncontingent (response independent) reinforcement produced significantly faster response deceleration than nonreinforcement. Affective response did not differ between nonreinforcement and noncontingent reinforcement conditions.

Contingency awareness, partial reinforcement effects, and curiosity motivation in children

Third- and fourth-grade children were given a two-choice discrimination learning task, designed to associate each of three syllables with a different reward schedule: 100% reward, 50% reward, and 0% reward. Subsequent to the conditioning phase, measures were made of the subject's awareness of the reward contingencies associated with the syllables, as well as the pleasantness and curiosity which these syllables had acquired during the conditioning. The results showed that a partially rewarded cue was evaluated as somewhat less “pleasant” than a continuously rewarded cue, but evoked more “curiosity” than either the continuous or nonrewarded cues. These effects emerged only for the contingency aware groups.

Two-Choice Discrimination Learning in Children as a Joint Function of Incentive Level and Punishment

Abstract Sixty-five fourth- and fifth-grade boys were run in a two-choice discrimination learning task for punishment, high or low value reinforcement, or the combination of punishment with high or low value reinforcement. A significant main effect of reinforcement revealed that only Ss in the punishment group showed significant evidence of learning. The results were discussed as possibly being due to the distraction caused by tangible reinforcement; although, an alternative explanation was offered on the basis of previous findings that males perform better for punishment than for reward.

The effect of partially valid cues on children's discrimination learning

Second grade children were administered a two-choice discrimination learning task in which irrelevant dimensions were correlated .50, .75, or 1.00 with the 100 percent rewarded cue. The results indicated that when a cue was partially correlated with reward .75, learning was retarded. When all cues were relevant and redundant ( r = 1.00), learning was most rapid. These results supported an implication of Hypothesis Testing Theory that partially correlated or partially valid cues would not be eliminated rapidly, thus retarding acquisition of the correct response.

Visual discrimination learning of random figure problems by rhesus monkeys

In a two-choice discrimination learning paradigm, a U-shaped function was found between trials to criterion and the number of sides comprising the random figure discriminanda. These data are consistent with results from human Ss and suggest similar information processing capabilities in the two species. Changing the orientation of the stimulus figures after the discrimination had been learned only partially disrupted subsequent discrimination performance. Any single difference in selected form attributes between the two figures comprising a problem was not a good predictor of learning speed, suggesting that monkeys use a combination of such factors.

Two-choice discrimination learning in children as a function of punishment modality and reinforcement combination

In an effort to replicate and extend a previous finding that there are no differences between reinforcement and punishment in discrimination learning with children, 160 third grade boys were run on a two-choice discrimination learning task which compared the effects of punishment presented in the same and in different sensory modalities as the reinforcement. The training conditions were reinforcement (token) for correct choices, punishment (tone or response cost) for incorrect choices, and reinforcement-punishment combinations (token-tone or token-response cost) for correct and incorrect choices respectively. The major finding was that punishment presented in the same sensory modality as the reinforcement was superior to punishment presented in a different sensory modality from the reinforcement.

Visual acuity and discrimination learning in the dolphin (Lagenorhynchus obliquidens)

This experiment investigated the underwater visual acuity of one adult female Pacific white-sided dolphin (Lagenorhynchus obliquidens) by assessing its ability to discriminate between stimuli comprising one and two vertical lines presented simultaneously in a two-choice discrimination learning apparatus. The experiment was performed in two phases: an initial training phase in which a basic two-line vs. one-line discrimination was established, and a testing phase which utilized this discrimination learning in determination of visual acuity. Visual discrimination learning throughout the training phase, was unexpectedly slow and apparently highly problem-specific, despite the use of a sequence of problems designed to facilitate initial learning. During the testing phase of the study, eight two-line vs. one-line test problems, differing only in two-line gap widths were presented in a random sequence 100 times each during 10 daily sessions, at a minimal viewing distance of 45 cm. A standard psychophysical difference threshold criterion was applied to the performance data, yielding an estimate of underwater visual acuity of approximately 6′ of arc, a value roughly comparable to that for the cat in air. The results of the study are interpreted as suggesting that dolphin vision is probably sufficiently well-developed for it to be a behaviorally significant modality, but that vision may well play only a relatively minor or specialised role in the guidance of normal underwater behavior.

The effects of irrelevant dimensions and stimuli on two-choice discrimination learning by children

Eight 9-year-old children were presented 24 eight-trial simultaneous two-choice visual discrimination problems. Each problem contained eight different stimuli. In half of the problems the stimuli varied along two dimensions with one dimension irrelevant, and for the other 12 problems the stimuli varied along three dimensions with two dimensions irrelevant. Even though the number of irrelevant stimuli was the same, the two-irrelevant-dimension problems were significantly more difficult than the one-irrelevant-dimension problems. There was significant learning on both types of problems and no significant interactions.

Note on the effects of a gross-motor approach to training attention control on discrimination learning in mentally retarded SS.

It was hypothesized that a gross-motor approach to training (walking board) attention control would generalize (à la Zeaman and House) to a two-choice discrimination learning task. Results for 22 Ss support this hypothesis.

Two-choice discrimination learning as a function of stimulus and event probabilities.

Two-choice discrimination learning as a function of stimulus and event probabilities.

Incentive effects upon attention in children's discrimination learning

Following a 40-trial training procedure in which high and low incentive values were conditioned to stimulus dimensions, 276 children in Grades 2, 4, 5, and 6 were tested on an 80-trial, two-choice discrimination learning test. Differential, incentive associated, dimension preferences from training were hypothesized to facilitate or impair test performance by altering observing responses to the relevant dimension in which the correct stimulus eue was found. After exploring boundary conditions of instructions, incentive type. definition of samples as learners or nonlearners, and mode of stimulus presentation in training, the hypothesis was clearly confirmed for males in the second grade. Equivocal support was found for girls and for older children under varying experimental conditions. Results, taken together with back-ward learning curves, strongly suggest that differential incentive values influence the point at which learning starts, rather than slope or asymptoie. The application of attention theory under certain conditions was commended as a possible explanation for some past findings of acquisition differences as a function of different reinforcement values.