Please enable JavaScript.
Coggle requires JavaScript to display documents.
Stimulus Control and Concept Learning (Behavioral Contrast (--fig 9-5,…
Stimulus Control and Concept Learning
GENERALIZATION GRADIENTS
Measuring Generalization Gradients
Probe trials
are those in which the other stimulus are briefly presented to measure the subjects responding but no reinforcer is given
Humans are able to undergo additional probe trials. For example, they can listen to the duration of a tone and respond if it is a given length or shorter or longer.
What Causes Generalization Gradients?
Pavlov stated that generalization is an automatic by-product of conditioning (innate)
Lashley and Wade stated that generalization cannot be obtained without some explicit discrimination training (learned)
Nondifferential training
is that in which all trials are the same
Presence-absence training
includes two phases: (1) Nondifferential training (S+), and (2) trials without the stimulus from the nondifferential training (S-)
Intradimensional training
is that which has stimulus of similar qualities but slightly different. Only one is reinforced.
Lashley and Wades hypothesis is supported. Nondifferential training produced a flat line. Presence-absence training produced a normal distribution. Intradimensional training produced a narrow normal distribution
Is Stimulus Control Absolute or Relational?
Spence's Theory of Excitatory and Inhibitory Gradients
Spence's experiments accounted for peak shift and why responding is so low to novel stimuli
You must subtract inhibitory strength from excitatory strength to understand the net associative strength
The Intermediate-Size Problem
The intermediate size problem favored relational. The net associative strength gave a sharper gradient but did not show peak shift.
Transposition and Peak Shift
Kohler's was based upon
transposition
, the relational belief would transfer to new stimuli with similar properties
and Hansen's was based off of
peak shift** which refers to a shift in the generalization gradient away from S- stimulus
Peak shift supports the relational position
Other Data, and Some Conclusions
Because of similarity, other S+ stimuli looked like the smaller of two choices, but the relational rule is to pick the larger of the two. Subjects picked the larger reinforcing the relational theory
There is strong evidence that animals are capable of relational responding
Absolute theory of stimulus control
states that a subject will remember all stimulus presented and any rewards they have and
relational theory of stimulus control
states that the subject will learn about the relationship between the stimuli and will choose the one that has greater incentive
Behavioral Contrast
--fig 9-5
Gutman
explored behavioral contrast in his experiment where subjects were given a multiple schedule. When one of the schedules was extinguished, responding decreased but increased in the other. This is referred to as
behavioral contrast.
This is an example of positive contrast and negative is when one of the schedules recieves additional reinforcement which then diminishes responing in the other schedules
A reason for positive contrast is that the subject has more time to do things during the extinguished sessions so that it can focus on responding during the working schedule
Another reason for positive contrast is that once the subject learned that more reinforcers could be received from the schedule they increased responding because they were trying to achieve more satiation
It is often impossible to predict how one stimulus will affect ones behaviors unless we also take into account other stimuli
It is dangerous to study reinforcement schedules as if they were separate entities. The subject is likely impacted by the others as well
Transfer of Learning and Learning Sets
Animals higher on the phylogenetic scale obtain learning sets faster than their lower counterparts. There are exceptions like rats with smells
After switching S+ to S- and S- to S+ after 11 trials, monkeys were able to recognize the change and adapt their learning set
A
learning set
is the improvement in rate of learning across a series of discrimination problems
Concept Learning
Concept learning is ambiguous with categorization
The Structure of Natural Categories
Three characteristics of natural categories are (1) that people tend to agree about which examples are central and which are peripheral, (2) people list central instances more frequently when questioned about a category, and (3) people take longer to decide if peripheral are a part of a category
When determining centralness,
exemplar
means to compare the new instance to memories of many individual examples and
prototype
means an idea is formed from experiences and if the new instance is similar to the prototype idea
Central instances
are members that fit well in a category and
peripheral instances
are members that do not fit well in a category.
Animal Studies on Natural Concept Learning
With trees as S+ and other as S- on slides,
Herrnstein
was able to get chickens to discriminate between the two
Human and animal categorization is the same.
Developing Stimulus Equivalence
Stimulus equivalence
refers to a situation in which subjects learn to responses to all stimuli in a category as if they are interchangeable
--fig 9-7
Equivalence sets and learning language are similar.
"Errorless" Discrimination Learning
Ideally, this errorless procedure could be introduced into education meaning errors would be much lower frequency on average
A drawback from this method is that subjects may have a difficult time learning discriminating reversals, where S+ and S- are reversed
Instead of introducing the extinction later, show it early on and only for a brief period
Stimulus Control in Behavior Modification
Stimulus Equivalence Training
Being used to teach those struggling to read. They learn equivalences between hearing a word, seeing the word and reading the word. Used in complex college course, trained students averaged 97% compared to their 57% counterparts
Study Habits and Health Habits
If you restrict undesirable behavioral activities, you will do them less... duh
Insomnia
If you do a lot of things in bed besides sleep, then it may become difficult to sleep in your bed. Eliminate excess activities where they don't belong to properly associate other things like bed and sleep.