Lecture 3: Thorndike Flashcards
Thorndike, general
Systematically investigated the influence of rewards and punishment on learning and problem solving: The first of the true learning theorists
Thorndike is the first motivation or drive based theorist
Teachers college
o Therefore, he had a profound effect on education in the U.S through his students
Originally interested in animal research
o but then later applied his findings to humans
Believed fundamental laws of learning hold true despite difference among species
o This position has been supported in the literature
Species differences are largely in the degree of what can be learned,
There doesn’t seem to be difference in how things are learned
It’s the degree of learning not how of it
• This is the original S-R learning
But S-R in an operant sense
o Believed learning was based on association between sensory impressions (S) and impulses to actions (R )
Although he was considered an early behaviorist,
all the things he talks about goes on inside the organism
o Sensory impression and impulses are all internal
o Neural firing/impulses
o Afferent (S) and efferent (R) nerves = responses in S-R association
Thorndike’s Research, approach and methods
Most of the subjects were cats, dogs and chickens
Studied these subjects using his own creation, a puzzle box
A box that distributed food by pulling lever, or that could be opened by pulling/pushing another lever/pole, or by stepping on a specific platform
He was interested in the quality of the actions and the speed in which the animal was able to perform the proper response = Latency
Started timing as soon as the animal was placed in the box
• He measured this on a learning curve
o First person to really consider this
o These results are based on groups of cats not individual cats
o Curves thereby represented mean data
He rejected the concept of insight
Learning, especially motor behavior, occurs only through trial and error
If a cat watched another cat do it, it didn’t help them
Even when he did show them how to do it, it didn’t help
The only important thing was experience
The current response is obtained by chance accident, the food strengthens the response to occur more and more
Law of effect is derived from this
He called reinforcers “satisfiers” and “annoyers”
Thordike’s 3 Laws
Law of Effect
Law of Readiness
Law of Exercise
Law of Effect
When a connection between a stimulus and response is made (S-R) and this is followed by a satisfying state of affairs (S+) the connection is strengthened.
If the connection is followed by an annoying state of affairs (S-) the connection is weakened
Bond between stimulus and responses are stamped in by satisfiers and stamped out by annoyers
If something is to be learned a satisfier/award need to be presented
For something to be eliminated a punisher must be present
Research on Law of Effect
Over the years, he did a number of studies to revise the law of effect
Using a list of pronouns, and simple “yes” and “no” responses, the results showed that rewards increased response and thus the law worked
When they presented annoyers then the response did not decrease, thus he concluded that punishment did not work→ punishers did not reduce the frequency of response (this conclusion is wrong because they were not annoyers)
He revise the law of effect
o when a series of reinforcers is followed by a satisfying state of affairs, the connection is strengthened
o only rewards worked
o dropped the negative side of the law of effect
• punishment did not necessary weaken the connections
He believed punishment can affect learning indirectly by leading the subject to do something else, and that something else would lead to reward and learning
Punishment = something that weakened the connection between a stimulus and response
Thereby, decreasing the probability of that response when the stimulus occurred
However, based on his definition, what he did by saying “no,” “wrong,” was not a satisfying punishment technique
o however due to his influence, the idea that punishment doesn’t work became the dominant theory in psychology and parenting
Skinner and Estes later concluded similar findings
However, in this case, they failed to use proper punishment
In all of these cases, they only examined a few punishment techniques
In order to properly judge punishment, need to test as many punishment techniques as one can think of
Punishment is the correct response for undesired behaviors and it works quickly but you have to use it sparsely because
It leads to habituation which would lead to a loss in effect
Leads the organism to becoming aggressive, possibly vengeful
Law of Readiness
A preparedness for making the next response in a sequence must exist if learning is to occur
Failure to fulfill or to be ready = annoying
Being allowed to fulfill = satisfying
- not biological readiness
e. g. if forced to eat when one is not food deprived violates the law of readiness
Law of Exercise
Connections are strengthened with use and weakened with disuse
Strengthening = an increase in the probability that the response would be made when the situation occurs
Without feedback, performance did not improve no matter how many repetitions occurs
Revised version: mere repetition is not enough, practice must result in success or reward
Practice is not a value per se, but only because it permits reward to act upon connections, e.g. praise
o this makes the law of exercise a corollary of law of effect
Theory of Intelligence
Intelligence is a matter of how many specific S-R bonds you have
The more specific S-R bonds, the more intelligence
Learning = forming more S-R bonds
o hence the name bond theory, or hyphen theory
There is no G, no insight, and no sudden understanding
Transfer of Training
Allows new learning to occur without having to rely on a the start of a completely new process
How much generalization you get depends on how many similarities are in the original learning
Learning, however is specific, not general
o if it seems general, it is because they are many identical elements between the new and old situations
Positive transfer: similar elements make it more easy to learn something new
Negative transfer: similar elements make it more difficult to learn something new
Spread of Effect
the effects of reward spread to adjacent S-R bonds
a biological effect that happened in the brain [“a bleeding-effect”]
spread of energy across the connections in the brain
in reality, this didn’t occur in his studies because he did them wrong
Hull, Skinner and Thorndike
Hull and Skinner followed the approach of Thorndike
Hull believed himself to be a more scientific, objectified version, of Thorndike
Hull did not object to internal constructs but believed there was a way to measure them
Skinner aimed to focus more on the external, measurable, variables
e.g.you don’t say something is hungry, but rather state that an animal has not eaten in x-hours
This makes the internal variable external, allowing for one to measure it
Watson’s approach was Pavlovian and was really bothered by Thorndike’s work
Watson ridiculed Thorndike and his focus on internal constructs because they are unmeasurable
o i.e. you can’t measure annoyers and satisfiers
