Instrumental Conditioning

study 1

given either sucrose or food, some apired with LiCl to see if reduces lever pressing (makes sick)

P groups should show less responding

sucrose/LiCl groups do press lever - food pellet/LiCl do press sometimes - sucrose was always left cus sticky

supports Tolman's cog account - have encoding of consequence of actions, but animals in LiCl still making response so law of effect must still be working too

study 2

group 1 = 500 training trials - group 2 = 100 trials

group 1 had no differences in response to outcome - always press lever

group 2 differed - if outcome was devalued then reduced number of presses (when paired with LiCl)

when have more trials, they dont learn devalued outcome - both SR and cog account going on - instrumental is mediated by both systems

lesioned infralimbic and prelimbic cortex

2 training regimes:

1. lever press 1 in context 1 with 15 sessions (highly trained response and became habitual)

2. 5 sessions in different context with different reward

they then changed value using sensory specific satiety (devaluation)

condition 1 should be habitual so they keep pressing lever

condition 2 still association between behaviour and outcome so press less

sham lesion: low training (less response to devalue) and high training (no difference in response) - when lots of training dont understand devalue

prelimbic region lesioned animals not able to show sensitivity to actions and always habituate (similar response behaviour is mediated through prelimbic)

infralimbic lesioned animals still show behaviour and reduce action as result of devalue - goal directed behaviour

inactivation of infralimbic cortex by injecting muscimol creates a temporary method

used same method as kilcross and coutureau 2003

working cortex: no effect of devaluation in overtrained/habitual animals

inactive cortex: allows animal to behave according to own goals

systems work in parallel

amphetamine causes OA increase and sensitisation - so need less amphetamine to get behavioural effects of drug and sensitisation increases learning

1. amphetamine --> learn lever press --> food

2. learn lever --> food --> amphetamine

experiment 1

does drug abuse change experience of actions? -  devaluation

if no amphetamine there was less response to devaluation (goal-directed) - if have amphetamine there was no difference in response (habitual) - they therefore form a habit because of sensitisation

experiment 2

amphetamine after training meant less response to devaluation - (need amphetamine before training to produce a stimulus response/habitual behaviour)

conclusion:

pre-training sensitisation with amph produces SR habit behaviour --- post-training sensitisation doesnt retrospectively change the goal directed system

is prelimbic cortex involved in learning or expressing goal directed learning

comapred lesions before and after conditioning

sham lesions showed goal directed behaviour

lesion --> training = habitual

training --> lesion = goal-directed

prelimbic cortex has time limited role - necessary for forming association but dont need pre-cortex to display appropriate behaviour when its been learnt

learning vs. drive tested on hungry animals

1. given food reinforcer (learn food is good)

2. half not

then all have access to food and lever press training when not hungry - dont learn condition

test to press lever to get food when hungry or not - learn food is good when hungry so should show less response when full (unless realise all food is good when hungry - primary motivation)

hungry and exposed to lever - show lots of repsonse so must have learnt to know food will help hunger

instrumental only sensitive to learning

pavlovian sensitive to primary motivation