Effects Of Overtraining Research Articles

EFFECT OF OVERTRAINING ON REVERSAL AND NONREVERSAL SHIFTS IN NURSERY SCHOOL CHILDREN.

A previous investigation (5) has shown that nursery school children, like nonverbal organisms, learn a nonreversal shift faster than a reversal shift. This experiment was designed to find out if overtraining which is known to facilitate a reversal shift in infrahuman subjects (8) would have a similar effect on young children. The results indicate that overtraining had a facilitating effect on reversal shift and no significant effect on a nonreversal shift. This was interpreted as support for a mediational explanation of the effect of overtraining. However, it may be that the mediators are of a perceptual type previously used to explain similar effects in infrahuman subjects (2, 6) rather than mediation by implicit verbal labeling (4, 5).

EFFECTS OF OVERTRAINING ON REVERSAL AND NONREVERSAL DISCRIMINATION SHIFTS IN A FREE OPERANT SITUATION.

Rats were trained on a single-stimulus, successive discrimination in a free operant situation. An irrelevant stimulus dimension was present at ail times. Following attainment of the acquisition criterion, Ss were shifted immediately or given 4 or 8 days of overtraining before being shifted. Half of Ss were given a reversal shift and half a nonreversal shift. Overtraining did not affect reversal or nonreversal learning. These results were contrasted with those of Mackintosh (1962). Various theoretical issues were discussed in light of the present findings. It was tentatively concluded that the overtraining effect depends upon the role of observing behavior in the formation and overtraining of a discrimination.

Effects of Overtraining on Acquisition and Reversal of Place and Response Learning

Five groups of rats were given 0, 25, 50, 75, and 100 overtraining trials in a T-maze to determine the effects on the acquisition and reversal of place and response habits. Increasing overtraining produced more response than place learners, but there was no facilitative effect in reversal learning with increased overtraining, nor was there a differential effect on the ease of reversal of place or response learning.

Overtraining and Transfer within and between Dimensions in the Rat

Although reversal learning in the rat is normally assisted by overtraining, it seems that transfer to a new discrimination in a different stimulus dimension is not. This effect is confirmed in the present experiment, even though the stimuli of the second problem had been present (and necessarily attended to) in the first problem. This result is taken to show that the effects of overtraining cannot be accounted for by supposing that it strengthens an overt orienting response of the animal. A second group of rats was run to show that transfer to a second discrimination problem within the same dimension as the first is facilitated by overtraining. The important variable, therefore, that determines the direction of the effect of overtraining on transfer between problems, is the relationship between the stimulus dimensions concerned.

Further Studies of Overlearning and Position Reversal Learning

The effects of extensive overtraining on position reversal were studied in 2 experiments in which the stimulus complex associated with the choice response was manipulated. In Exp. I a visual cue (brightness) was correlated with the position response. In Exp. II the stimulus consequences of an incorrect response were made highly discriminable by associating an abrupt illumination change with an incorrect choice. Despite these modifications the overlearning reversal effect failed to occur in both experiments, verifying a number of previous reports.

THE EFFECT OF IRRELEVANT CUES ON REVERSAL LEARNING IN THE RAT

Twenty‐four rats were trained on a brightness discrimination, given 0 or 150 overtraining trials, and then reversed. For one group only one irrelevant cue (position) was present throughout the experiment. For the other two groups the reversal was learnt with orientation as a second irrelevant cue; in one case this cue had been present during original learning, in the other it had not. Where there were two irrelevant cues, overtraining was shown to have a greater facilitating effect on reversal than where there was only one—whether or not the second cue had been present in original learning.The results are interpreted as suggesting that the effect of overtraining is to increase the probability that responses will be controlled during reversal by the relevant stimulus dimension.

Comment on “Effect of Overtraining on Subsequent Learning of Incidental CUES”

Church (3) has recently reported a set of experiments designed to test the conditions under which animals, while making discriminations on the basis of one set of cues, will pick up information about newly introduced incidental cues that later may be used to guide discriminative choice when the original cues are removed from the situation. Like Blum and Blum (1) and Bruner, Matter, and Papanek (2), he found that rats do indeed learn something about such incidental cues while operating on previously established cue-discrimination habits. The heart of the issue about which I wish to comment briefly has to do with a hypothesis introduced by Bruner, Matter, and Papanek (2 ) and tested by Church (3 ). It is that if the animal is highly overtrained in his response to the original cue to which he is responding and is highly motivated, there will be less likelihood of his picking up the incidental cue when it is introduced into the situation. By highly motivated, we can mean the 36-hr. food deprivation used by Bruner, et nl. or the 22-hr. water deprivation used by Church. Overlearning of a cue is a matter of 200 to 300 trials of practice with the original cue before introduction of the incidental cue which can also be used to guide behavior in the situation. The general hypothesis proposed by Bruner, et nl. was that when behavior becomes mechanized with overlearning, the organism ceases to elaborate cue-goal cognitive maps or expectancies. Church in his ingenious experiments introduces an interesting and worthwhile distinction concerning incidental cues of the type under discussion. The distinction is between new cues that are introduced and previously irrelevant cues that are made relevant at a given point in the progress of overlearning. Bruner, et ul. used the latter procedure, Church the former. In the first of these experiments, rats were responding to black or white doors in a four unit T-maze, with position of the doors randomized. After a certain amount of overlearning, the correct doors were always arranged in an LRLR or RLRL sequence and. finally, the color cue was removed and animals tested for their mastery of single alternation. In short, a previously irrelevant cue suddenly became relevant and we ask under what conditions the animals '.notice3' that the cue has become relevant. In Church's first experiment, a newr cue was introduced after a period of overlearning. Rats went either to one or I!?? other arm of a single-unit T-maze, depending on whether a black or n hit? card cue was present on the side. The incidental cue added after over

Effect of Overtraining on Subsequent Learning of Incidental Cues

Effect of Overtraining on Subsequent Learning of Incidental Cues

Effects of overtraining in conflict.

Effects of overtraining in conflict.