The most striking difference between man and lower animals is the ability to communicate with symbols. The most obvious example of such behavior is the use of language. Partially because of the importance of language development, psychologists have studied verbal behavior extensively.
This unit traces the steps in verbal learning, including acquisition, forgetting, recall, and the factors that influence them. We will begin with acquisition. As you read through the text, ask yourself the following questions.
It is difficult to conduct a learning experiment that is not influenced by what the subjects already know
In an early effort to find how verbal associations were formed and retained, Hermann Ebbinghaus in 1885 initiated a series of experiments on human memory. He was soon confronted with the problem of selecting items which were not already meaningful to the subjects so that previous learning would not contaminate the results. As a partial solution, he developed the "nonsense syllable." These short "words" made up of three-letter combinations were arranged in a consonant-vowel-consonant pattern, such as PIV or DAX. They were carefully selected so that only those with no definite meaning in ordinary language were used. Psychologists still employ the nonsense syllable as an experimental tool, but we now know that even nonsense syllables have some degree of meaningfulness in that they can evoke associations.
Nonsense syllables are frequently used in verbal learning experiments. The tasks in such studies usually require students to learn to associate the nonsense syllable with something else, an abstract form, a symbol, or another nonsense syllable. This activity is referred to as paired-associate learning. A typical experiment follows.
The subject's task is to learn the following stimulus-response (S-R) pairs:
S -- R
The subject learns that when shown the stimulus @ he must respond by
saying "BIX." Usually this is taught by a procedure called the method of
anticipation. The subject may be given the following instructions:
"First, you will be shown a figure like this. @
"Then, shortly thereafter, you will be shown the figure with a word written below it. @ BIX
"Your task is to learn to anticipate the word by saying it aloud before it is shown.
"You will be shown the entire list over and over again until you can go through it twice with no errors.
Each repetition is called a trial block. The procedure is continued until a criterion of mastery is reached, such as two consecutive blocks with no errors. The experimenter records whether or not the subject correctly anticipates each response. The data are frequently summarized graphically.
DISTRIBUTION OF PRACTICE
An experimenter must define exactly the learning that is expected to take place
Some of the earliest experiments in paired-associate learning were performed to answer the question, "Is it better to practice a task continually until one is proficient, or is it better to space the practice over a longer period of time?" For example, if a soldier were given 50 rounds of ammunition for practice, would he learn better if he shot 3 rounds, then rested; shot another 3, rested; and so on -- or would his performance be superior if he shot all 50 rounds in one practice session? The first procedure would be called distributed practice and the latter, massed practice. Some of the variables to be considered are:
A study was conducted by Lorge (1930) to investigate various rest periods. The task required the subject to learn to trace a complex pattern while viewing his movements in a mirror.
The results of this and similar studies indicate that rest periods do improve learning, but there seems to be an optimal length of rest period. With any further) increase in rest lengths, there is little additional gain.
KNOWLEDGE OF RESULTS
The soldier's target practice would have shown no improvement at all if he had no way of knowing how far from the bull's-eye each shot was. He must have knowledge of the result of each shot in order to benefit from the practice.
Feedback to the learner is a critical event in most learning experiences
In a paired-associate learning task, the presentation of the S-R pair after each response tells the subject whether the response was correct, and what the correct response should be. The experimenter is thus providing the subject with immediate knowledge of results. Knowledge of results or "feedback" has a significant effect on many kinds of trial and error learning tasks. The old cliche, "we learn from our errors," is true only when we have knowledge of the outcome of our acts. Knowledge of results is effective for at least three possible reasons. It may:
An experiment by Baher and Young (1966) demonstrated the role of feedback in a simple task. The task required the subject to draw a 4-inch line while blindfolded. The subjects could feel a 4-inch piece of wood which served as a reference at any time. Knowledge as to the correctness (+ .2 inch) of responses was provided only half the subjects after each trial. The other subjects received no feedback. The difference between the two groups for the first eight days reveals that the feedback group got 60% right after 8 days while the non feedback group got nly 12% right.
(Fig. 3 shows a striking difference, where the feedback group got 60% right after eight days; while the non-feedback group got only 12% right).
Knowledge of results (like reinforcement) must be given immediately to be effective. Delay in providing feedback lessens its effect. If the task requires a subject to associate a response with the outcome of that response, feedback must occur prior to his next response. This is why a lecture course with infrequent tests yields lower levels of learning.
THE EFFECT OF MEANINGFULNESS ON ACQUISITION
Some nonsense syllables are easier to learn than others. Although nonsense syllables are deliberately selected to avoid meaning in ordinary language, they do, in fact, produce meaningful associations. A syllable that resembles a real word has some of the associative strength of that word. Therefore, nonsense syllables such as KOP, SEK, and REK are more easily recalled than, say, MIP, KIV and SEF. Several methods have been used to scale the meaningfulness of nonsense syllables. All these methods involve some measure of association. Generally, subjects are presented with a stimulus item, such as PIV, and asked to think of an association to that word. If five out of fifteen subjects give associations to the nonsense word PIV, while ten subjects provide associations to DAX, the experimenter could conclude that DAX is more meaningful than PIV. This same technique can obviously be used with almost any stimulus item, including real words. A few real words, such as URN, are less meaningful by these measures than are some "nonsense" syllables!
Meaningfulness of material does affect its ease of acquisition. It must be carefully noted, however, that meaningfulness as defined here refers to associative strength. It does not indicate whether the material is "relevant" in the subject's life or deeply "felt" by him.
Now test yourself without looking back.
1. In an experiment by Gagne and Baker (1950) subjects learned to identify various colored lights by labeling them with letters of the alphabet.
What is this type of learning called?
2. In the Gagne and Baker experiment described above, would you expect the fastest learning between the first and second trial blocks or between the third and fourth trial blocks?
3. If subjects in the Gagne and Baker experiment were being taught using the method of anticipation, the
response required of the subjects during the learning task would be to:
a. say nothing, but carefully observe each S-R combination as it was presented.
b. say the appropriate letter aloud when a particular colored light was turned on.
c. place a card with the appropriate letter beneath each of the colored lights.
d. (none of these)
4. Match each example of feedback with the major function that it probably serves.
1. Provides additional information
2. Acts as incentive
3. Signals remediation
a) "That was a good job, Johnny."________
b) "You are weak at adding fractions." _______
c) "You are shooting three inches too high."_______
d) "Nice backhand you have there."________
e) "No, the answer is 1066."________
"I'm afraid you're not good enough yet to make the team."________
5. An industrial psychologist wishes to design a training course to teach eight separate (not related) operations; and so would probably:
a. have the workers learn all eight operations in one sitting.
b. allow three-minute breaks between each learning task.
c. allow no break between each learning task.
6. Which of these words would most likely be used in an experiment where it is important to eliminate the
effects of previous learning?
7. Here are some of the response members from two lists of S-R pairs:
If you wish to determine the meaningfulness of each of these items you would be most likely to:
a. ask the subjects to define each.
b. ask the subjects to think of as many associations to each word as they can.
c. ask them to rate the relevance of each word on a ten point scale.
8. Suppose after scaling the above words, we find that items in List I are more meaningful than those in List II If we now design a learning experiment with two groups of subjects, Group I learning List I and Group II learning List II as responses to the same stimuli, which group (Group I or Group II) would be most likely to achieve mastery first?
MODULE 1 EXERCISES
Ellis, Muller, and Tosti (1966) scaled the meaningfulness of random shapes using several different measures of association. In one condition the subjects were shown the shapes one at a time and asked to give as many associations to each shape as possible in a one minute interval.
The results for five of these shapes are shown below:
|Average Number of Associations|
Which shape was the most meaningful? _______________________________________________ 3
We know that learning and recall are facilitated by meaningfulness. The shapes in the experiment described above were later used in a learning-recognition task. Which would you predict?
In the method of anticipation, the subject's task is to anticipate the
response to a stimulus just prior to the presentation of the
response and stimulus together. Which of these learning methods
comes closest to the method of anticipation?
a. Children reciting the multiplication tables in unison
b. A child responding aloud to a flash card with problems on one side and answers on the other
c. (neither) ______________________________________________________ 5
Knowledge of results is the feedback that a person receives
about the correctness of his response. Check the statement(s)
that provide knowledge of results.
a. "Your answer should have been George Washington, not Abraham Lincoln."
b. "That was a job well done!"
c. "I've got a stomach ache."
d. (none of these)
Feedback is an important factor in many kinds of learning tasks. In a very large manufacturing plant, the inspection department notified the foreman that all the widgets made by a new man were 2mm too large. What should the foreman do to facilitate this man's job training? _______________________________________________________________2
1 a, b
2 givefeedback; tell him his widgets are too large.
3 shape D
NOW TAKE PROGRESS CHECK 2
1. A child learns by rote to say "four" when presented with the stimulus "2 + 2." This is an example of:
a. reflex learning.
b. paired-associate learning.
c. concept learning.
d. distributed learning.
2. Which of these learning curves are most typically found in acquisition of verbal S-R pairs?
(graphs left out)
3. Sam and Bill, both novices at shooting, practiced with 100 rounds of ammunition. Sam rested after each shots, but Bill shot continuously until his entire 100 rounds were spent. The curves below show the average number of bull's-eyes for each block of 10 shots.
(graph left out)
4. A teacher designs a device to teach children the names of animals. First a picture appears, then the same picture with the name below it. What is the method being employed here?
5. A psychologist presented nonsense syllables to subjects and asked them to think of as many
associations to each syllable as they could. The mean number of associations given were then used as a
measure of that syllable's:
d. (none of these)
6. A teacher presents addition problems on flash cards. To the card which shows 2 + 2, a student responds "five." The teacher then says, "No, it's four." What is the teacher providing?
Click here to return to Table of Contents Module 4
Go on to Module 2