Learning And Motivation Final

Front 1

generalization

Back 1

responding similarly to similar stimuli

Front 2

Guttman and Kalish

Back 2

Trained pigeons to peck at one color

Test phase: showed a variety of colors, some close to training stimulus, some not

Results: birds pecked more at colors that were more similar to the training stimulus

Conclusion: Birds' response gradient is a function of how similar test stimulus was to training stimulus

Front 3

Steepness of response gradient

Back 3

precise measure of stimulus control

Front 4

Flat response gradient

Back 4

org responds in similar fashion to all test stimuli, stimulus feature varied among test stimuli does NOT control behavior

Front 5

Steep response gradient

Back 5

org responds more to some test stimuli than to others, stimulus feature varied among test stimuli DOES control behavior

Front 6

Riccio CTA Study

Back 6

Gave rats LiCl paired with chocolate milk, split into two groups
GI: Tested 24 hours later, would drink sweet milk but not chocolate milk

GII: Tested 1 week later, avoided both sweet milk and chocolate milk

Conclusion: GII forgot specific stimulus attributes, and thus generalized their aversion to include sweet milk

Front 7

Discrimination training

Back 7

teaches organisms not to generalize

Ex) Two keys, S+ and S-, learn that if peck S+ get a reward, no reward if peck S-

Ex) To test if birds can see color, reward for one color but not another, if go back and choose the rewarded stimuli, can see color

Front 8

Jenkins and Harrison

Back 8

Discrimination training in pigeons

GP I: 1000 cps tone (S+), no tone (S-)
GP II: 1000 cps tone (S+), 950 cps (S-)
GP III (control): 1000 cps tone for all

Results:
GP I: Peak responding at S+
GP II: Tighter responding, animals learned that small difference in sound=reward, least generalization of all groups
GP III:----
Conclusion:
control group=intermediate stimulus control
both training groups=more stimulus control than control group, less generalization

Front 9

Two conclusions of discrimination training

Back 9

1) increases stimulus control of instrumental behavior, decreases generalization

2) particular stimulus (eg. tone frequency is most likely to gain control of responding if the S+/- vary along that dimension

Front 10

Intradimensional discrimination

Back 10

S+/S- vary only in terms of the value of the stimulus feature

a form of expert performance

Front 11

Peak shift (only occurs with intradimensional discrimination training)

Back 11

the shift of the peak of the generalization gradient in the OPPOSITE direction of

Front 12

Hanson (1959)

Back 12

(S+)=a 550 nm light
All groups had the same S+, diff S-

GP 1:( S-)=590 nm

GP 2: (S-)=555 nm

GP 3: (S-)=550 nm

Results:
GP1: Peak response is closest to 535 nm
GP2: Peak responding far away from S+, around 530 nm

Conclusion: Degree of peak shift depends on similarity btw S+ and S-. The more similar S+ is to S-, the greater the degree of peak shift.

Front 13

Spence's theory of peak shift

Back 13

There is an excitatory gradient that peaks at S+ and there is an inhibitory gradient that peaks at S-; therefore, the more similar S+ and S- are, the greater the degree of overlap.

Front 14

Discrete stimulus

Back 14

stimulus with a clear onset and offset

Front 15

Contextual stimuli

Back 15

also control responding

Ex) Rats and shock/food in room A/B

Ex) Pigeons get food in noisy context, no food in quiet context, become very active when put in noisy chamber

noisy chamber=S+
quiet chamber=S-

Front 16

Honig (1963)

Back 16

line discrimination study with pigeons
-Took animals into context 1, where S+ is key with vertical line and S- is key with horizontal line
-In context 2, S+ is horizontal line, S- is vertical line
-Animals had to use context to learn what "rules of the game" are
-Test=put animals in room with a bunch of different cues they could peck at (lines at diff orientations), animals could learn that when in certain context, rules were like they were in training

Conclusion: S+/S- were discrete stimuli, context=modulator

Front 17

modulator

Back 17

stimulus that signals a relationship between between two other events, used when talking about instrumental conditioning

Front 18

occasion setter

Back 18

usually related to CC where context signals relationship btw CS and US

Front 19

Two basic procedures used to study aversive control of behavior

Back 19

1) avoidance procedures
2) punishment procedures

Front 20

avoidance procedures AKA active avoidance

Back 20

subject has to make an instrumental response to avoid/prevent something bad from happing

Ex) kid cleans room to avoid being punishment

Ex) study to avoid doing badly on a test

Increases instrumental response

Front 21

punishment procedures AKA passive avoidance

Back 21

Instrumental response brings about a consequence/outcome

Ex) passive avoidance lab task

Decreases instrumental response

Front 22

Bechterev (1913)

Back 22

had people put finger on metal disk
warning tone/light
shocked finger periodically

avoidance response=operant conditioning, removal of finger changes the outcome (does not happen in CC eg. rat freezing (CR) does not change fact that gets shocked)

Front 23

Brogden study

Back 23

2 groups of guinea pigs in running wheel
-Gave them CC-->CS (tone) + US (shock)
UR=run on running wheel

GP1 (CC)=shock always presented 2 secs after tone came on, did not matter whether guinea pig ran or not (beh has no effect on outcome)

GP 2 (Avoidance conditioning)=if guinea pig ran and moved wheel during tone, avoided getting shocked

Results:
GP1=increase running a little, then decrease, no very strong conditional response
GP2=within a few days, avoiding shock 100% of the time

Conclusion: Avoidance is different from CC because results from two groups are not identical

Front 24

Discriminated/Signaled Avoidance Task

Back 24

each trial initiated with CS, if do not make response, CS (tone) followed by US (shock)

early trials=escape
later trials=avoidance

Front 25

2-way shuttle avoidance

Back 25

rat jumps back and forth

If animal crosses quickly=avoids shock
If animal does not cross=gets shock

Front 26

1-way shuttle avoidance

Back 26

have to pick animal up and put back

animal only learns to jump one way

Front 27

Two-Process theory of avoidance

Back 27

assumes that 2 mechanisms involved in avoidance learning
1) Classical conditioning of fear to some CS
Ex) 2 way shuttle avoidance-->learn tone paired with shock, tone becomes aversive

2) Instrumental negative reinforcement of the avoidance response (eg. shuttle response) through fear reduction due to termination of CS

Ex) shuttling on later trials reinforced by turning off of CS because CS has become aversive (due to its being paired with shock)

Ex) Studying for test reduces anxiety caused by fear of failure

Front 28

Experimental testing of the two-process theory

Back 28

Acquired drive experiments

Front 29

Acquired drive experiments

Back 29

1) pure CC of fear to CS-->animal's beh has no effect on outcome, just a CS-US pairing (tone and shock)

2) subjects can shuttle to turn off tone (CS), subjects not shocked so shock not motivating behavior

Results: animals increased shuttling to turn off CS