Background Transfer of learning is the effect that prior learning has on later learning. Transfer of Learning ThorndikeLocke It is the fundamental goal.

Background Transfer of learning is the effect that prior

learning has on later learning.

Transfer of Learning

Thorndike Locke

It is the fundamental goal of education: to make the skills and information students acquire in the classroom helpful to them outside the classroom.In 1700, the British empiricist philosopher, John Locke, proposed a theory of transfer called The Doctrine of Formal Discipline. It was challenged two centuries later by American psychologist, Edward L. Thorndike, with his Theory of Identical Elements. Thorndike founded educational psychology.

Doctrine of Formal Discipline

Locke maintained that transfer depends on the amount of effort you put into mastering a task, not the specific skills or information you acquire. The mind is like a muscle. You have to exercise it to make it stronger.


Thorndike Locke

Here is the theory in Locke’s own words, with a few alterations to update the phrasing:

Doctrine of Formal Discipline

“Would you have a man reason well, you must [give him practice ahead of time], exercising his mind in observing the connection of ideas and following them in train.


Thorndike

“Nothing does this better than mathematics, which therefore I think should be taught to all those who have the time and opportunity, not so much to make them mathematicians as to make them reasonable creatures......that having got the way of reasoning, which that study necessarily brings the mind to, they might be able to transfer it to other parts of knowledge as they shall have occasion.”

Theory of Identical Elements

Thorndike maintained that transfer takes place to the extent that the original task is similar to the transfer task.


Locke

It depends on how how many “elements” the two tasks have in common.

For example, taking a high school course in geometry may help you later in life if you become a surveyor or navigator, but it won’t help you if you become a lawyer. You won’t strengthen a general ability to think logically by taking geometry.

Thorndike’s Theory of Learning Thorndike’s transfer theory was

based on his general theory of learning that said learning consists of forming connections between specific stimuli and specific actions (S-R learning).


Locke

He did pioneering experiments on animal learning in which he studied how cats learned to escape from a small box (a “problem box”) by stepping on a pedal or pulling on a piece of string. It was the forerunner of the Skinner box.

Thorndike’s Theory of Learning


Locke

The cats learned gradually apparently through blind trial-and-error. Responses that opened the door were repeated; responses that didn’t open the door eventually stopped occurring.

Thorndike formulated the original version of the Law of Effect: Responses that lead to positive outcomes are “stamped in”. Responses that lead to negative outcomes are “stamped out”.

The result of this process is an S-R bond.

Studying Transfer

There are two kinds of transfer, positive and negative. In positive transfer, learning on the original task speeds up learning on the transfer task. In negative transfer, the effect is the opposite: Learning on the transfer task is slowed down.


Thorndike Locke

In a typical experiment, there is an Experimental Group that gets both tasks, and a Control Group that gets only the transfer task. The researcher compares the two groups’ rates of learning on the transfer task. Faster learning by the Experimental Group indicates positive transfer; slower learning indicates negative transfer.

Studying Transfer


Thorndike Locke

Experimental Design

Experimental Group

Control Group

Learn Original Task

Rest

Learn TransferTask

Learn TransferTask

Phase 1 Phase 2

Studying Transfer


Thorndike Locke

Experimental Group

Control Group

Learn Original Task

Rest

Learn TransferTask

Learn TransferTask

Phase 1 Phase 2

Suppose the groups took the following numbers of trials to learn the tasks:

15 8

5

The Experimental Group showed __________ transfer.

negative

Studying Transfer

It’s negative transfer because the Experimental Group took longer to learn the transfer task than the Control Group. The number of trials to learn the original task is irrelevant. Ignore it, even though the Experimental Group shows a decrease on the transfer task. Maybe Task 2 was easier than Task 1. Just compare Experimental and Control on Task 2.


Thorndike Locke

What about the following set of numbers...?

Studying Transfer


Thorndike Locke

Experimental Group

Control Group

Learn Original Task

Rest

Learn TransferTask

Learn TransferTask

Phase 1 Phase 2

8 14

19

The Experimental Group showed __________ transfer.

positive

Paired-Associate Learning

Thorndike’s theory of transfer talked about the number of elements two tasks have in common but did not distinguish between stimulus elements and response elements. Transfer may depend on which elements are involved.


Thorndike Locke

To investigate the role of stimulus and response elements, researchers have presented paired-associate lists as the original and transfer tasks. The tasks can then be made similar in terms of their stimuli, their responses, or both.


The following notation system is used to refer to the paired-associate lists in Phases 1 and 2. There are two pairs of letters, like A-B, C-D. The first letter of a pair represents stimulus items in the list and the second letter represents response items. The above notation means that both stimuli and responses in the two lists are different.


Thorndike Locke

Suppose the two lists used the same stimulus items but different response items. The first list would be represented as A-B. The second would be __ –C.A



Thorndike Locke

Experimental Group

Control Group

Learn A-B

bok - xiz

Rest

Learn C-D

foh - taw

Learn C-D

foh - taw

Phase 1 Phase 2

A-B, C-D

Typical Result: Positive Transfer


Even though the two lists are completely different, the Experimental Group often outperforms the Control Group on the transfer task. Findings like this have led to a distinction between specific transfer and general transfer.


Thorndike Locke

A-B, C-D

Specific transfer depends on similarity between stimulus or response elements of the information presented. General transfer does not; it occurs for other reasons.


It is likely that the Experimental Group benefits from experience in Phase 1 because they learn how to memorize paired-associate lists under the conditions of the experiment. Such general positive transfer has been called learning-how-to-learn: learning something useful about a type of task or problem that can be applied to any instance of it.


Thorndike Locke

A-B, C-D

Thorndike’s theory can handle this because it says that task elements can be either part of a procedure or part of the substance of what’s learned. The effect is more limited than Locke would expect.



Thorndike Locke

Experimental Group

Control Group

Learn A-B

bok - xiz

Rest

Learn A-C

bok - taw

Learn A-C

bok - taw

Phase 1 Phase 2

A-B, A-C

Typical Result: Negative Transfer


This negative transfer goes against Thorndike’s theory. Half the elements (the stimuli) are identical in the two lists so there should be some positive transfer. Negative transfer occurs because you have to extinguish (stop saying) List 1 responses before you can learn the List 2 responses.


Thorndike Locke

A-B, A-C

It’s like learning to drive on the left side of the road after many years of driving on the right side. There will be a strong tendency to keep driving on the right.



Thorndike Locke

Experimental Group

Control Group

Learn A-B

bok - xiz

Rest

Learn A’-B

bik - xiz

Learn A’-B

bik - xiz

Phase 1 Phase 2

A-B, A’-B

Typical Result: Positive Transfer

(similar stimuli, same responses)


This positive transfer is readily explained in terms of a principle of operant conditioning that says when you learn a response to a stimulus, you will tend to make that same response to similar stimuli. This is the principle of ...


Thorndike Locke

A-B, A’-B

It helps to explain why learning one “romance” language, like French, can help you learn another, like Italian or Spanish...

Stimulus Generalization



Thorndike Locke

A - BFrench: bon good

A’ - BItalian: buono good

Spanish: bueno good

Generally, the findings on transfer support Thorndike more than Locke. Transfer depends on details of the information being learned or the type of task involved.

Learning Sets

A “learning set” is a tendency to approach a problem in a certain way as a result of previous experience with similar problems. The term was introduced by Harry Harlow to characterize the remarkable improvements he observed in monkeys’ performance on discrimination problems as a result of their being given hundreds of problems of the same type.


Thorndike Locke

Review of Terms

Two other terms were discussed that basically mean the same thing as learning set. They are...?

Learning Set = ?

Learning-How-to-Learn =


Thorndike Locke

General Positive Transfer

Learning Sets

In Harlow’s experiments, monkeys were presented with two objects that differed in multiple ways, like shape, color, and size; for example, a red cylinder and a blue pyramid:


Thorndike Locke

If the monkey pushed aside one of the objects (here, the red cylinder), he found food. If he chose the pyramid, he got no food. Left-right positions were reversed randomly from trial to trial.

Learning Sets

After the monkey consistently made correct choices, Harlow switched to a new set of objects, like a green cube and a silver sphere:


Thorndike Locke

The monkey got trial after trial until he always made correct choices. Then Harlow switched to a new pair of objects and the animal had learn a new solution. Altogether, there were 344 problems, each with a different set of objects.

Learning Sets

The monkeys became “experts” at solving this type of problem. The first few problems took a lot of trials to solve—blind trial-and-error like Thorndike’s cats in the problem box.


Thorndike Locke

After 300 problems (not trials on the same problem), they solved each problem within 2 trials, the absolute minimum, using a “win-stay, lose-shift” strategy.

If the first object they chose was correct, the chose it on every trial. If it was wrong, they shifted to the other object on Trial 2, and then stuck with it.

Learning Sets


Thorndike Locke

1 2 6

Trials

100

75

50

Perc

en

t C

orr

ect

R

esp

onse

s

Problems 1 - 8

Problems 33 - 132

Problems 289 - 344

These general improvements in problem-solving support Locke, but they’re specific to the type of problem the animals practiced, which supports Thorndike.