learning and behaviour ch 10
This preview shows page 1. to view the full 5 pages of the document.
Stimulus Control and Concept Formation
•Stimulus control- relationship between stimuli and beh.
•Measuring generalization gradients
○A pigeon is reinforced to peck at a yellow key for food
○Now we wish to measure generalization with a different colour key
Probe trail- briefly present other colour to measure responding but do not
Or begin with the same type of training and follow with a series of extinction
trials with both yellow and other colours.
Children were given a 40 second tone, then they were asked to say yes if it
was the same tone or no if it was a different tone. There were more yes’ to
tones that were about 4 seconds and more no’s if they were shorter or
•What cause Generalization Gradients?
○Pavlov- generalization is an automatic byproduct of the conditioning process-
○Lashley and Wade- Generalization gradients are not innate, they depend of
learning experience. Subjects must receive experience in which reinforcers occur
in the presence of one or more colors but not in their absence.
•How experience affects shape of generalization gradients
○Jerkins and Harrison conducted an experiment that provides support for Lashley
and wade’s theory
Two groups of pigeons responded on a VI schedule with a 1000Hz tone.
Three pigeons received non-differential training, in which every trial was
the same, light was lit, tone was on and VI schedule in effect.
Five other pigeons received presence- absence training, which included
two types of trials. (1) 1000Hz tone that were exactly as the other group,
and (2) without tone, which the key light was lit as usual but no reinforcers
They were all given extinction trials in which different frequency sounds
The non-differential training group remained about the same no matter the
sound, and the presence-absence training group, typical gradients with
sharp peaks at 1000Hz.
○Another experiment by Jerkins and Harrison
Two pigeons received discrimination training in which 1000Hz was an S+
and a 950Hz was an S-
Intradimensional training- discrimination training in which S+ and S- come
from the same stimulus continuum (sound frequency)
Most responding occurred around 1000Hz and no much when Hz was 950 or
less, or more than 1010Hz.
○Thus, generalization gradient are dependent on experience. With no differential
training tone frequency has no control over responding, with presence-absence
training, tone frequency exerted modest control.
You're Reading a Preview
Unlock to view full version
Only page 1 are available for preview. Some parts have been intentionally blurred.
○Other studies showed that you generalization gradient can occur without
experience, but Lashley and Wade suggested that animals might have learned to
pay attention to certain things from nature.
•How Sensory Deprivation Affects the Shape of Generalization Gradients
○Some generalization gradients require experience and some do not
○With birds sensory deprivation showed that wavelength of light can occur even
when a subject has had no previous exposure to other colour, when it comes to
tones birds need experience.
Is Stimulus Control Absolute or Relational?
•Stimulus discrimination procedure- two stimuli are presented together and the
subject must choose between them. (Dark grey or light grey)
•Absolute Theory of Stimuli Control-the animal has learned about the two stimuli
separately. 1 gives me food, the other doesn’t.
•Relational Theory of Stimulus Control- the animal has learned about the
relationship between the stimuli.
Some theorists believed that with non human subjects it was absolute, because non
humans could not understand relationship
•Transposition and Peak Shift
○ (Kohler)Transposition- the subject has transferred the relational rile to a new
pair of stimuli. Ex. A dark and light grey – choose light grey, a light and very light
grey, choose very light grey.
○(Hanson)Peak shift- similar to transposition, but found in general gradient
○Successive discrimination procedure- the positive and negative stimuli are
presented at different times
○With training a peak occurred earlier than without training
•Spence’s Theory of Excitatory and Inhibitory Gradients
○A clever version of absolute theory
○An excitatory gradient develops around S+ and an inhibitory gradient develops
○Associative strength- ability of stimuli to elicit response
○Spence decided that’s the strength of any stimuli can be determined by
subtracting, its inhibitory strength from its excitatory strength
○Makes more sense when looking at the graph on pg. 231
•The Intermediate-Size Problem
○Supports Relational Theory
○Chimps were given stimuli of 3 out of 9 squares all different sizes, 1 was smallest
area and 9 the largest.
○Example: If they were given 1, 4,9 they were supposed to pick the middle size.
○Results: they chose the middle one most of the time, no matter which ones were
○Therefore they were judging based on relationships between the squares
•Evaluating The Two theories
○ Absolute is best when stimuli are not presented together and relational is best
when stimuli are presented together
•a phenomenon in which responding in the presence of one stimulus changes as a result of
a change in the reinforcement conditions during another stimulus
You're Reading a Preview
Unlock to view full version