LEARNING & MEMORY

LEARNING & MEMORY

1. Learning is subject to cue competition effects e.g. overshadowing, blocking etc

2. PREDICTION ERROR

   • When the CS is presented, nothing is predicted, and the US occurs. The learning systems adapts to minimise and eradicate prediction error.

3. HOW DOES PREDICTION ERROR LOOK IN THE BRAIN?

   • THE REWARD SYSTEM
     • OLDS & MILNER (1954)
       • Implanted electrodes in different brain regions in rats and electrically stimulated regions when a lever was pressed. Found that stimulation of midbrain regions caused rats to repeatedly press the lever to the extent where they would not eat or drink for extended periods of time.
         • This shows the function of rewards in learning and a physical correlate for reward in the brain
     • Activation of midbrain dopamine neurons in rats when the US is delivered can potentiate learning (Steinberg et al, 2013) – we can make a reward more rewarding when this brain area is stimulated (we can exaggerate it).
       • Activation also attenuates blocking presumably because it reinstates the prediction error.

4. INFORMATION-PROCESSING MEMORY MODEL

   • ATKINSON & SHIFFRIN (1971)
     • A stimulus is presented, and a representation of that stimuli becomes active in sensory memory. That representation moves from sensory memory to STM. Items in active in STM can produce responses or reactions to things that have just happened. This memory either decays or goes into LTM.

5. DOES THE CS RETRIEVE THE US FROM MEMORY?

   • COLWILL (1994)
     • One group of rats, during the first stage of the experiment, were presented with a tone and then received food. In the second stage they received food and lithium chloride (which makes them sick). The other group were presented with a light and then received sucrose and in the second stage, were given sucrose. In the test phase, group 1 were presented with a tone and group 2 was presented with the light.
       • Results showed that rats showed significantly less food searching activity in the presence of the tone and increased food searching activity in the presence of the light. This suggests the CS is retrieving the US from memory, such that they know that the tone is associated with food that makes them sick.
   • A CS-US association permits the CS to facilitate the retrieval of the US from LTM. Learning occurs when there is no representation of stimuli in the STM at the time in which they occur. Priming reduces surprise and therefore learning. Priming puts stimuli in the STM before it is presented and made consciously aware of.

6. There are 2 types of priming in which a stimulus can enter the STM

   • 1) Self-generated priming
     • A stimulus primes itself upon presentation
       • TERRY (1976
         • Demonstrated that proximal US presentations decreased speed of learning. This suggests that US is in the STM because it has just been presented, it is primed so less learning occurs.
     • Self-generated priming of the CS
       • BEST (1979)
         • Group one receives a flavour and then lithium chloride – makes them feel sick. Group two receive a flavour and then later, receive a flavour and then lithium chloride. Results showed that group 1 subsequently drank less in a test trial because they have learned that the flavour makes them ill whereas group 2 drank more. This suggests that they had a weaker association between the flavour and illness, and this was due to self-generated priming.
   • 2) Retrieval generated priming
     • Also known as associative priming from LTM
       • The blocking effect (Kamin, 1969)
         • There was more learning in the control group because there was no retrieval of the shock from memory. It was not primed and thus, more learning occurred about the light and the shock. Blocking is consistent with the attenuation of learning by retrieval-generated priming.
     • Retrieval-generated priming of the CS
       • LATENT INHIBITION
         • KAYE & PEARCE (1987)
           • Learning was slower in the familiar group due to previous exposure to the light in earlier training. This is an example of retrieval-generated priming because the pre-exposure to the light is happening on a different day than the conditioning trials. So, they have learned that the context predicts the light and when they are put back into the context, the context retrieves memory about the light and they therefore don't learn about it very well.
             • Long delays between exposure and tests – latent inhibition is likely a consequence of retrieval-generated priming. Changing the context reduces latent inhibition (Hall & Channell, 1985).
   • An important conceptual point is that, in the real world, we don't have CSs and USs, we just have stimuli. They don’t differ. They are just learned associations between two stimuli. So, priming should decrease learning not just about the USs but also CSs?

7. PUZZLE:

   • If something is present in STM, we don’t learn about it very well e.g. priming. BUT, sometimes, when something is in STM it supports learning.
     • So, is there more than one type of STM?

8. WAGNER (1981) SOP model

   • The central principle of SOP is that a stimulus representation is in one of three dynamic states
     • When the cat encounters the food, elements in the corresponding cat-food representation are activated. The course of activation is determined: there is a brief period of focal processing during which the elements are in a state designated A1. These elements do not stay long in this focal activation state, however, but decay passively into a secondary, more peripheral state designated A2. Eventually, when the food is no longer present, the elements passively decay further into inactivity (I)
       • Retrieval generated priming – When we retrieve something, it goes into from I to A2, A2 doesn't support learning and there is a slow decay process back to I.
       • Self-generated priming – Representation is inactive, stimuli is presented and then is activate into A1 but then decays very quickly into A2.