Discriminative Training On Two Different Luminance Of Key Lights Essay, Research Paper

Not a few experiments on the discriminative training with pigeons were done over past several decades, and many researchers found that various factors relate to the results of discriminative training with pigeons. In a classical study by Heinemann and Rudolph (1963), they suggested that the geometric size of the stimulus influences on the efficiency of the learning by pigeons. There are several other factors pertaining to that efficiency of the learning. According to Sargisson and White (2001), the training delay is one of those factors. They found that the longer the training delay, the more sessions were required for all birds to reach the same level of response accuracy (Sargisson & White, 2001).

In this discriminative training, therefore, the ring doves were trained with several different conditions in order to obtain some suggestions for to understand how they learn and what factors relates to the efficiency of their learning. What the doves learned in this operant conditioning was the discrimination of two different luminance of key-lights. Because many researches found that the doves had the reliable ability for discrimination of the intensity of lights, the results of learning depends on the long term memory of the subjects.

Summarily, the purpose of this training was to know whether the doves can discriminate the differences of luminance of lights, and the factors which influences on their learning.

Method

Subjects

Seven hungry ring doves (Streptopelia risoria) were tested. Six of them had previous experience on training to discriminate between the two red lights which had different luminosity. Before that, all of the doves were autoshaped to peck at a red key illuminated by one red bulb. Two of them have participated in many experiments for 14 years.

Apparatus

Testing was conducted in a student chamber (37 cm wide ? 29 cm deep ? 23 cm high) designed by Lehigh Valley. The side of the chamber was clear plastic board to see the reaction of the doves. Stimuli were the light produced by one or two red bulbs. The each bulb had approximately 120 cd/㎡of luminosity. The light produced by two bulbs had approximately 240 cd/㎡of luminosity. The size of stimuli was the circular switch which was approximately 2 cm in diameter. It was presented at the point of where 7 cm from the floor. The reinforcement was 2 seconds in the discrimination training. The food was mixed grains.

The seven chambers were located in the different sound attenuated cubicles. The chambers were operated not by computers but the hands of students. The number of pecks on the key-lights with sufficient force was automatically counted by the machine connected to the chamber. That machine controlled the luminosity of red lights. Basically, pecking the switch with one red light delivered reinforcements (opportunities to eat mixed grains for 2 seconds) to the subjects . No punishment was used in the training.

Procedure

Four types of conditionings were conducted in different days. The memory test was done at first. It had two parts. The first part of the test was conducted in order to make the birds remembered that pecking the red key would make food available. The second part was done to make birds remembered that the pecking produces the feeder when the key is red, not when it is dark.

The next conditioning was the first session of discrimination training. During this session, the subjects was rewarded with food for pecking only when the key light was illuminated with one red light (S+) and not rewarded when the key is illuminated with two red lights (S-). The subjects were given 60 trials (30 S+ and 30 S- trials). Then, the number of pecking during S- periods (30 seconds) was counted. The key light was turned off between each trial for an inter-trial interval (ITI) of 10 seconds.

The second session was the same as the first session of discrimination training except for that the number of S+ and S- trial were 15 each, and the reinforcement schedule was set up during S+ trials. 5 ratios were presented for each S+ trial as random without replacement.

The third session was exactly the same as the second session of discrimination training except that the stimulus control tests were given. In the stimulus control tests, S+ period was given in exactly the same way S- periods were given. That is, the S+ period lasted for 30 seconds with no food, and the number of pecks were counted.

Note that the number of S- trials for two birds in the first session and one bird in the second session were less than 30 or 15.

Results

Overall, the results of memory test were successful. Only one bird pecked on the dark key many times, but the others did not peck so many. It means that most birds could successfully remembered that the pecking on the dark key did not produce the feeder, and they could get food when the red key was illuminated.

In the first session of the discrimination training, the range of the mean number of pecks during S- trials for all birds was from 37 to 65. In the second session, it was from 8 to 73. One bird showed drastic reduction of the sum number of the pecks during S- trials. Although, the number of pecks of the bird, 4341 slightly increased, that bird had no previous training in discrimination between pecking keys illuminated with one or two red lights. Thus, the overall results clearly showed that most birds performed better. That is, they began to realize that the pecking on two red light-key did not produce the feeder.

In the third session, some birds performed better, and some other birds showed slightly worse performances. The number of pecks during the stimulus control test (S+ test) for all birds were higher than the number of pecks for S- trials. Then, the discrimination ratio of all birds was higher than 0.5.

The bird, 4557 performed very well. In the memory test, after 5th trial out of 30, he did not peck on the dark key at all. In all of three sessions of the discrimination training, he showed the amazing performances as shown in Figure 1. The number of pecks during S- period continually decreased, and the discrimination ratio was 0.93 which was the highest among all subjects.

Discussion

Discrimination ratio which were obtained after the three discrimination training were above 0.5 in all subjects. Therefore, it can be said that the doves could discriminate the differences of the luminance and learned what the one or two lights meant. Heinemann and Rudolph (1963) used the three different sizes of stimuli in the discrimination training and found that the smallest stimuli had the greatest efficiency for pigeon*s leaning. The smallest stimuli they used had an inner diameter of 1 inch which was very similar to the stimuli used in this conditioning. Thus, it was one reason of that all doves showed excellent discrimination ratio and even the subject, 4341 which had no previous experience showed the sign of leaning. And also, two red lights had twice the luminosity of one red light, so the overlapped area of the distribution curve of sensation for each light might have been small. These explanations tells about the advantageous conditions for this training.

However, the important purpose of this training was to get some ideas about how they learned. Paying attention to the bird, 4557 which had the highest discrimination ratio might give some valuable ideas for that. In the first session, he began to realize that pecking on two red lights did not produce the feeder. He pecked only 5 times in the last S- trial. However, the first S- trial in the second session, he pecked 23 times. It looks like a sign of that he did not learn in the first session, but it does not. In the second session, the procedure called random without replacement was used, so the largest number of pecks required to get food during S+ trials was 25 times. The fact that he pecked 23 times explains that he did not notice the differences of intensity of lights effects the availability of food at the beginning, but he remembered he had S- trial in the previous session and stopped pecking. That is why the number of pecking did not exceed 25. According to Sargisson and White (2001), the delay interval plays an important role in the efficiency of learning. The longer the delay interval, more chance of forgetting arise. Although their experiment was not a discrimination training, the connection should exist. The ITI was 10 seconds for all sessions of the discrimination training. The interesting point was as follows. In the second and third sessions, the bird, 4557 achieved several 0 pecking in S- trials, but each of those trials was always in between the S+ trials. If two or more S- trials continued, he had always pecked several times during those S- trials. In other words, he never consecutively achieved 0 pecking. It might explain that he forgot what he learned or at least confused during the consecutive S- trials which means that he had quite long time for refraining of pecking. Because S- trials lasted 30 seconds and the interval time between the trials was 10 seconds, he had to refrain pecking for 90 seconds or more if two S- trials were consecutively given. It strongly tested if he really learned or not. In short, he certainly discriminated the differences of the intensity of key lights, but the forgetting or confusion might have occurred.

In one of S- trial, he pecked 31 times, however this trial was in between the S- trials he pecked only 1 or 2 times. Thus, the reason why he pecked so many might be explained as a typeⅠerror in signal detection theory which inevitably occurs sometimes. And also, he was so hungry that it might have been an expression of his irritation. Do the warm-blooded animals have the feeling of irritation? Further researches are needed for that question, but the bird, 4557*s sudden intense pecking during S- trials in the second or third session might have been reasonably explained if the feeling of irritation were taken into consideration.

In conclusion, the high discrimination ratio and the many numbers of pecking during S+ test trials clearly showed that the bird, 4557 learned the discrimination of two different intensity of lights, and the other birds also successfully showed the sign of learning. However, most birds had previous experience, so this operant conditioning had an aspect of remembering. It might have been the reason why most birds had pretty high discrimination ratio. Therefore, it can be said that the results was reliable but the validity of this experiment might have been questionable. This training clearly showed the changes of the dove*s behavior but did not give clear factors for how they learned. That is to say, the possibility of that they just remembered what they had learned previously could not be rejected from this experiment.

Bibliography

Heinemann, E. G., & Rudolph, R. L., (1963). The effect of discriminative training on the gradient of stimulus-generalization. The American Journal of Psychology, 76, 653-658.

Sargisson, R. J., & White, K. G., (2001). Generalization of delayed matching to sample following training at different delays. Journal of the Experimental Analysis of Behavior, 75, 1-14.