Avoidance Learning Flashcards
Types of avoidance learning
Passive avoidance rat must withhold a response (‘don’t do this or else’)
Active avoidance rat must make a response (‘do this or else’)
Shuttle avoidance rat actively avoids, no one chamber is safe
The shuttlebox
Avoidance apparatus for rats and mice
Classical conditioning
Bolles 1971
●Species-specific defence reactions (SSDRs) e.g., flight and freezing (Bolles, 1971)
●By classical conditioning animals learn to respond to stimuli associated with pain
●SSDR of flight good basis for active avoidance, freezing for passive avoidance
●CRs like URs as expected on the basis of stimulus substitution
●Avoidance responses like the SSDR (e.g., running away) are much more easily learned
Model of fear and pain
Bolles and Fanselow 1980
Fear can potentiate a variety of SSDRs and can inhibit other motivational systems, including pain
The fear and pain systems are functionally organized to allow an animal to defend itself and recover from attack
All avoidance behaviours not the same: preparedness
the 3 possible avoidance responses, running was far easier to learn than turning or standing (in order to avoid receiving an electric shock in the running wheel)
Classical conditioning
Bolles 1971
Problems
- It is difficult, but possible, to train animals to perform avoidance responses that do not resemble any SSDR
- Avoidance learning is faster when the designated response causes omission of the aversive US than it is in a group exposed only to a classical contingency
- After just a couple of signalled shocks, an animal may respond so efficiently to the CS that it never gets another shock. Why doesn’t the conditioned fear extinguish?
Two Factor theory
Miller Mowrer 1960
●The rat continues to respond because he is reinforced by shock omission? But why should the rat continue to expect a shock?
●We’d need to include the idea (from Two-Process Learning Theory) that reducing exposure to the classically conditioned warning signals is also reinforcing
●In active avoidance, animals learn to escape the (secondarily aversive) stimuli that warn shock is imminent
●Adaptive anticipation: motivated to reduce fear, better to escape sight rather than fangs of tiger
The two factors in the two factory theory of avoidance
On early trials, before the animal knows how to avoid, there are trials in which warning signals are paired with the to-be-avoided shock
Because of Pavlovian fear conditioning, these stimuli become warning signals and arouse fear (making the response terminates the shock)
On later trials, when the organism has learned to make the response that avoids the shock, the response terminates the warning signal (as well as preventing the shock)
This allows reinforcement through fear reduction
Classic test of Miller Mowrer theory
Kamin 1956
Direct test in the shuttlebox
4 groups of rats (buzzer CS, shock US):
1)Standard escape-avoidance: response avoided or terminated shock and terminated buzzer – learned ?
2)US avoidance: Could not terminate buzzer but could avoid shock – learned ??
3)CS termination: Could terminate buzzer but could not avoid shock – learned ??
4)Classical conditioning: same buzzer-shock pairings but responding had no effect on exposure to buzzer or shock – ????
Results of classic test of Miller Mowrer theory
Kamin 1956
1) Learn √√√
2) Some learning (but CS-duration fixed at 5 seconds, so response was followed by CS termination with slight delay) √
3) Some learning (but shuttling was effectively punished) √
4) No instrumental learning X
Clinical implications
Depression
Learned helplessness
Seligman and Maier 1967
In Phase 1, shock was escapable (by making an instrumental response), inescapable delivered on a matched or ‘yoked’ schedule, or there was no shock
In Phase 2, when allowed to escape and/or avoid shock (in a different environment) the group exposed to inescapable shocks did poorly
Clinical implications
Anxiety
Many innate rather than the result of a learning experience
But behaviour therapy works and learning theory principles help with analysis…
Fear of snakes/dirt usually persists without any actual contact/infection
Because avoidance responses or rituals can act as CIs
Watson and Little Albert experiment
Study
According to two-process learning theory,
through classical conditioning stimuli gain reinforcing and motivational properties
Avoidance learning can be very persistent because
the execution of the avoidance response generates potential safety signals
Thus CIs protect conditioned fear from extinction
Protection from extinction
Soltysik et al 1964 1983
Stage 1:
Cats had pairings of mild shock with two CSs (CRs paw-flexion, changes in respiration and heart-rate)
CI was presented 2 seconds after the onset of CSa and CSb on trials when shock was omitted
CI came to inhibit CRs to both CSa and CSb
Stage 2:
Both CSa and CSb presented for 180 trials without shock but only CSa was followed by CI
Stage 3:
Test of conditioned fear to CSa and CSb (measured as CRs) when each was presented without shock or CI
Result:
Fear extinguished to CSb but not to CSa
Two Factor theory and conditioned inhibition
Avoidance example:
light -> shock
(light+avoidance response) -> no shock
So the avoidance response would be a conditioned inhibitor
But how does this help us?
Have we explained resistance to extinction?
Do conditioned inhibitors have any special properties that would help?
Appetitive example:
light -> food
(light + tone) -> no food
So the light is excitatory and tone inhibitory
In avoidance learning, making the response or its after-effects might act as a conditioned inhibitor
Classic test of Miller Mowrer theory
Kamin 1956
Kamin’s test of Miller-Mowrer Theory -
Shows very powerful effects of CS-termination on the learning of active avoidance
Problems:
•Again why doesn’t the conditioned fear just extinguish without reinforcement?
•Given the choice between signalled and unsignalled shock, animals nearly always choose signalled shock
A free operant avoidance procedure Sidman 1953
Brief electric shocks occur at regular (S-S) intervals unless a response is made which initiates a longer (R-S) interval
Notice that each response resets the R-S interval
Animals can make the response whenever they ‘want’ and thus learn to avoid shock despite the fact that there is no explicit warning signal
