Stimulus Control

Question 1

What is stimulus control?
Behaviour is said to be under stimulus control of a S+ when…

e.g., S+ example
phone ringing is DS for picking up the phone and speaking (consequence = entertainment or information)
i.e. behaviours you would not perform if the S+ was absent

e.g., S- Example
Someone reading the paper is a signal for you to not talk to them (consequence = nothing)

Answer 1

the behaviour is more likely in its presence than in its absence

S+ = SD (phone ringing) > Response (pick up and speak) > Consequence (info or entertainment)

S- = SD (reading paper) > Response (don’t talk) > Consequence (nothing)

Question 2

GENERALISATION
The tendency for the learned behaviour to occur in…
The transfer of a learned response from…

DISCRIMINATION
The tendency to behave differently in the presence…
Learning to respond to…

Answer 2

presence of stimuli that were not present during training
one stimulus to another, similar stimulus

of different stimuli
one stimulus and not another, similar stimulus

Question 3

GENERALISATION
Amount of generalisation is a function of…
e.g., got drunk on rum & developed taste aversion

Answer 3

how similar other stimuli are to the conditioned stimulus we learnt about
Rum is similar to spirits and taste aversion us transferred to spirts
Rum and beer are less similar > not much transfer

Question 4

Graphic representation of generalisation:
X-axis plots…
Y-axis shows the…

The slope of the line indicates the amount of generalisation:
The flatter the line…
The more peaked the line…

Answer 4

some dimension along which the test stimuli are varied (how similar things are to the original stimulus)
strength of conditioned responding to the different stimuli (strength of the CR)

the more generalisation (less discrimination)
the less generalisation (more discrimination)

Question 5

GENERALISATION GRADIENTS
Guttman & Kalish (1956)
Trained pigeons to respond to a single coloured key After acquisition tested the amount of responding to other key colours

Pigeons were reinforced on a VI schedule for pecking a response key illuminated by a yellow light (wavelength of 580 nm).
After training, the pigeons were tested with a variety of other colors presented in a random order without reinforcement, and the rate of responding in the presence of each color was recorded.

Results?

Answer 5

The pigeons had a high discrimination (low generalisation) for the different coloured keys and responded mostly to the coloured key in which they had been trained to press.

Question 6

THINGS THAT AFFECT THE GENERALISATION GRADIENT
1. Physical…
2. Perceptual…
3. Past…
Hearst et al. (1964) - examined range of lean (e.g. VI240s) to rich (e.g. VI30s) schedules during initial training on subsequent shape of generalisation function

Answer 6

similarity between conditioned & test stimuli (e.g., hose pipe and snake)
abilities and salience
experience (i.e. prior learning history) (e.g., the effect of reinforcement schedule during training)
when use richer reinforcer = greater generalisation curve

Question 7

DISCRIMINATION
Discrimination is all about…
We use the stimuli that are available to us as a…

Answer 7

responding it leads to reinforcement and not responding when it does not
guide for what is and what is not appropriate behaviour

Question 8

WAYS TO STUDY DISCRIMINATION:
(1) MULTIPLE SCHEDULES

e. g., VI 240s (reinforced once every 2 mins) vs VI 30s (reinforced once every 30s)
e. g., VI 15s (red) vs VI 300s (green) - light sometimes red or green (randomly alternate but spend same amount of time)

Answer 8

2 (or more) reinforcement schedules
Each schedule is associated with a different stimulus

VI 240s - lower rate of reinforcement
VI 30s - higher response rate per min

Test to see if pigeons can discriminate

• when red light expect pigeon to peck at fast rate
• when green light expect pigeon to peck every so often

Question 9

AN INTERESTING PHENOMENON: CONTRAST
Refers to the…
Multiple schedules; different S+ (lights) to distinguish them:

VI 30s (red) and VI 30s (green) > green = extinction

Positive contrast =
Negative contrast =

Answer 9

negative correlation between the response rates in the two components of a multiple schedule

Rate of response decreases in green, rate of response increases in red = positive contrast

rate of response increases in an unchanged component, with a decrease in behavior in the altered component.
rate of response decreases in the unchanged component with increases in response rate in the altered component.

Question 10

THINGS THAT INFLUENCE DISCRIMINATION PERFORMANCE

Answer 10

1) BIOLOGICAL COMPATIBILITY
e. g. rats have trouble discriminating colours

2) Stimulus Similarity
e. g. relatively easy to discriminate red from green

3) PAST EXPERIENCE
e. g. trouble discriminating between authentic vs forged paintings unless you’re an art expert

4) PRESENCE OF OTHER STIMULI
too many stimuli makes it difficult to discriminate which is the correct one

Question 11

Categories and Concepts
Transfer of learning about stimuli may result in…
Categories are groups of…

Answer 11

the formation of categories and concepts

stimuli or events that belong together

Question 12

Concept Formation
In a natural environment, animals must make difficult but important discriminations between objects.
1) Ability to respond differently to…
2) Once an animal has demonstrated it can treat different stimuli as members of different categories and…

Answer 12

different stimulus items indicates an animal can effectively categorise those stimuli.
assign novel stimuli to appropriate categories, then it is displaying an ability to form concepts.

Question 13

What happens if they are presented with two novel stimuli?

Answer 13

Animal has moved from being able to form categories to being able to form concepts

Question 14

From Category to Concept: Alex the Parrot
Alex learned to…
Alex was demonstrating…

Answer 14

categorise novel stimuli (e.g., shape, colour, and ‘matter’)

the ability to use a concept.

Question 15

Recap: Difference between Categories and Concepts
‘Knowledge’ or a learning history which results…
If a concept is in operation, novel stimuli or events (a sequence of stimuli) may be…

Answer 15

in the formation of stimuli or events into categories (or classes)
assigned to pre-established categories

Question 16

Animals & Perceptual Concepts
Herrnstein (1979)
projected…
If a slide contained tree it was an…
If the slide didn’t contain any tree it was an…
In generalisation tests using novel slides, the pigeons were able to…

Criticism:
didn’t learn the category ‘tree’, just the particular slides they saw?
something in common across all slides (e.g. green leaves)?

Watanabe et al. (1995)
Showed evidence of concept formation by training bird to respond to…

Answer 16

80 picture slides into an operant chamber.
S+ & pecking was reinforced.
S- & pecking was not reinforced.
discriminate between ‘tree’ and ‘not a tree’.

Monet (realist) and Picasso (abstract) paintings.

Question 17

EDUCATIONAL AND CLINICAL SETTINGS: Insomnia

For some individuals with insomnia, their bed doesn’t reliably occasion sleep (poor stimulus control)
Their bed may instead set the occasion for other behaviours (ruminating, reading, studying, browsing the internet)

Turner & Ascher (1979): Participants with Insomnia
Only go to bed when sleepy; do not read, watch TV or eat in bed; if you don’t fall asleep in 10 minutes, get up and do something else, only returning when sleepy; set alarm for the same time every morning; no day napping
Results?

Answer 17

CS (bed) > CR (sleep) > wake up feeling refreshed (consequence)
Insomnia: CS (bed) > CR (worry) > wake up feeling tired

Treatment reduced sleep onset to under 30 minutes

Question 18

EDUCATIONAL AND CLINICAL SETTINGS:
Errorless discrimination learning
A progressive way of training a conditional discrimination that results…
S+ = Red key presented for 30 s periods
S- = Started off dark and only available for brief 5s period - gradually became longer & greener over trials
Results?

Answer 18

in very few errors (i.e. responses to S-)

Fewer trials to reach accurate performance in ‘errorless’ group compared to ‘trial and error’ approach

Question 19

EDUCATIONAL AND CLINICAL SETTINGS:
Errorless discrimination learning
A progressive way of training a conditional discrimination that results…

Terrace (1963)
S+ = Red key presented for 30 s periods
S- = Started off dark and only available for brief 5s period - gradually became longer & greener over trials
Results?

Duffy & Wishart (1987)
Trained developmentally delayed students to discriminate shapes…
Progressing to recognising rectangle in the presence of other shapes
Results?

Answer 19

in very few errors (i.e. responses to S-)

Fewer trials to reach accurate performance in ‘errorless’ group compared to ‘trial and error’ approach

using errorless discrimination procedure
students made very few errors!

Question 20

Sometimes we have to be aware of other extenuating factors or stimuli:

Answer 20

Contrast effects and Stimulus factors

Question 21

Positive Contrast -
e.g. putting behaviour X in situation A into extinction leads to more behaviour X in situation B

Negative Contrast -
e.g. providing more reinforcement for behaviour X in situation A leads to less behaviour X in situation B

Meichenbaum et al. (1968)
Performed an intervention with troubled teenage girls:
Very disruptive classroom behaviour at baseline
Given money during afternoon classes for appropriate behaviour (changed component)
Not given money in morning classes for appropriate behaviour (unchanged component)
Results?

Answer 21

Responding increases in the unchanged component
Responding decreases in the unchanged component

afternoon class (changed component), morning class (unchanged component)
increase in behaviour in afternoon class (+ve contrast)
decrease in behaviour in morning class (-ve contrast)