Chapter 4 - Classical Conditioning Methods

Question 1

What are some explanations of Pavlovian conditioning?

Answer 1

1. conditioning results in activation of “US centre” in the brain (representation of the US) by the CS= S-S learning
2. conditioning establishes a new stimulus-response connection between the CS and the CR =S-R learning
   (option 1 - S-S Learning: CS substitutes for US; option 2 - S-R Learning: a connection is formed between the CS and the new CR)

Question 2

What evidence is out there that supports S-S Learning?

Answer 2

If the CS serves as a substitute for the US then the nature of the conditioned response should be determined by the US
prediction: CS’s conditioned with different US’s should elicit different types of conditioned responses (ex. autoshaping with grain vs. water- Jenkins and Moore, 1973 - light elicited two different responses)

Question 3

Describe the study performed by Timberlake and Grant (1975) on the nature of the conditioned stimulus?

Answer 3

• rat paired with presentation of another rat + food

- conditional responses of a participant rat was dependent on the CS (therefore, the US was forgotten)

Question 4

Describe the experiment performed by Rescorla (1975) in determining whether S-S Learning or S-R Learning was occurring?

Answer 4

Controlled experiment to test this: devalue the US
group 1: light = food (give lots of food)
group 2: light = food (restrict food) normal response to US

** the US still matters**

Question 5

Describe how learning can occur without a US?

Answer 5

1. higher order conditioning
   a) first order conditioning (CS1—> US)
   ex. donut store + donuts = happy
   donut store = happy
   traffic light + donut store = happy
   traffic light = happy (but weaker response)
   b) second order conditioning (CS2—>CS1)
2. sensory preconditioning: we learn associations between CS1 and CS2 right away

Question 6

What are the implications of learning without a US?

Answer 6

1. S-S associations may occur in the absence of a UR

2. Evidence for latent learning

Question 7

What four factors influence the strength and time required for classical conditioning?

Answer 7

1. # of pairings of the CS (ex. bell) and the US (ex. meat)
   • more pairings = stronger CR
2. the intensity of the US
   - ex. a squeak vs. a roar (more intense = faster conditioning)
3. how reliably the CS predicts the US
   - the more reliable the better
4. the temporal relationship between the CS and the US
   - less time = better (best at 0.5 seconds)

Question 8

What things make for an effective CS and US?

Answer 8

1. initial response to the stimuli
   - CS does not initially elicit the conditioned response
   - US elicits target response without prior training
2. significance and discriminability
   a) intensity/salience
   - give more > make the US more relevant to the biological needs of the organism
   - make it naturalistic > make the CS more similar to the kinds of stimuli that an animal will encounter in its natural environment
   b) novelty of the stimuli
   - CS pre-exposure effect (latent learning - delaying learning because a previous association has been learned)
   - US pre-exposure effect (doesn’t allow association to form)
3. CS-US relevance (belongingness)

Question 9

Describe the study performed by Garcia and Koelling (1966) on belongingness?

Answer 9

2 CSs: 1 audiovisual and saccharin
2 USs: 1 sickness or shock
test: audiovisual or saccharin (only one)
results: taste without audiovisual + sickness = more fear
audiovisual without taste aversion + shock = more fear
** more likely to condition things that belong together**

Question 10

What needs to happen for conditioning to occur?

Answer 10

The US needs to be surprising or unexpected for conditioning to occur

Question 11

Describe Kamin’s blocking study

Answer 11

group 1: tone –> US (paired)
group 2: tone –> no US (not paired)
conclusions:
1. temporal contiguity (pairing of a CS with a US) is not sufficient for Pavlovian conditioning
2. for learning to occur, the US must be surprising

Question 12

What is the Rescorla-Wagner model?

Answer 12

A mathematical model of Pavlovian conditioning

- surprisingness of the US determines how much conditioning will occur

Question 13

What are the assumptions of the Rescorla-Wagner model?

Answer 13

1. when a CS and a US are paired, an association is formed
2. The learning of this association is a curvilinear function
3. the effectiveness of the US in conditioning a CS is determined by how different the US is from what is expected
4. the expectation of the US is related to the conditioned properties of all the stimuli that precede the US

Question 14

Describe the Rescorla-Wagner model

Answer 14

1. associative strength of a CS increases with the conditioning until the CS perfectly predicts the US
   - associative strength of the CS/expectation of the US is denoted by V
2. US is denoted by lamda, thus, lamda is determined by the magnitude of the US
   - when the CS perfectly predicts the US then V is at its maximum (V=lamda)
   - lamda - V = amount of surprise (V approaches lamda with conditioning)
   - deltaV = lamda -V: V increases by deltaV until V=lamda
3. with each conditioning trial the US becomes less surprising, therefore the value of deltaV will decrease across trials
4. to quantify surprise

Question 15

What does the Rescorla-Wagner look like?

Answer 15

DeltaV = K(lamda-V)

- most learning occurs within the first experience

Question 16

What are the predictions the Rescorla-Wagner model makes?

Answer 16

1. conditioning (see notes for calculation)
2. blocking - ignoring other stimuli due to a previous perfect prediction
   - to unblock, we need to either increase or decrease the US (to change the expectation and therefore surprise the learner)
3. overexpectation - associative strength decreases because the expectation was not met
4. inhibition
   - similar to overexpectation
   - 2 types of trials: CS+ –> US and CS+–>CS-
   
   • trials where the US is present: Vcs+ = lamda
   • trials where the CS- is present: Vcs+ + Vcs- = 0
   • Vcs- must have negative associative value
5. extinction - the CS occurs without the US, leading to a decrease in CR
   - since there is no US during extinction, lamda = 0
   - the associative strength of the CS is therefore decreases from one conditioning trial to the next

Question 17

What are some problems with the Rescorla-Wagner model?

Answer 17

1. it doesn’t predict what happens to extinction for inhibitory associations

Question 18

What is the attentional model of conditioning?

Answer 18

attention to environmental cues that indicate the onset of a CS of a salient stimulus (US) critical for conditioning

Question 19

What are the three types of attention-based learning of CS-US?

Answer 19

1. looking for action: CS already predicts the US
2. looking for learning: US was surprising so looking for effective CS
3. looking for liking: US is salient so attend to any possible predictors

Question 20

What is the temporal coding hypothesis?

Answer 20

• takes into account the time between CS and US
• learn when (how long after) a US occurs in relation to a CS
• ITI/CS duration predicts magnitude of CR
  
  • relative waiting time hypothesis: how long the waiting time between CS-US compared to US-US
    
    • low ratio means CS barely predicts US better than previous US
• not very applicable