Memory 2

Memory 2

1. Serial Position Curve

   • Primary Effect
     • Tendency to remember the first items on a list
       • Interpreted as down to rehearsal
   • Recency Effect
     • Tendency to remember the recent items on a list
       • Interpreted as the capacity of the short-term store
   • Flat mid-part of curve
     • Interpreted as transfer to long-term store

2. Atkinson & Shiffrin (1968)

   • "Modal" model, or multistore model of memory, uses rehearsal to transfer to LTS
   • Rehearsal improves memory but it is neither necessary nor sufficient
     • Not necessary since the flat part of the curve is not zero even for unrehearsed items
       • Not sufficient since it doesnt always work
         • Glenberg, Smith & Green (1977)

3. Problems for a Unitary STS

   • Shallice & Warrington (1970)
     • Patients with only STM & only LTM but STM deficits aren't as devastating to LTM as we might expect
   • Baddeley & Hitch (1974)
     • Stimulate STM deficits by using tasks that should fill up the STS
       • Secondary tasks: sentence verification, semantic judgements, list learning, understanding of prose
         • All impaired by not devastated

4. Baddeley & Hitch (1974)

   • STS components
     • Central Executive
       • Attentional controller governing the flow of info to two slave systems
       • Visuo-spatial Scratchpad/Sketchpad
       • An Articulatory Loop/Phonological Loop
   • Baddeley's (1986) model of Working Memory

5. Visuo-Spatial Sketchpad

   • Brooks (1967)
     • Procedure
       • Visual
         • Pps asked to form a visual image of 4 x 4 matrix
           • Visual matrix when given instructions to imagine
             • Had to imagine the sequence of instructions using visual imagery
       • Verbal
         • Pps asked to remember instructions like the visual commands but made no sense
           • Has to remember the random words in the correct order
     • Findings
       • Pps remembered visual grid better than list of words.
         • 8 item spatial span vs 6 item verbal span
   • If you encode info in a visuospatial way, your memory is better than if you do it purely written

6. Dual Task Interference

   • Brooks (1968)
     • Spatial Task
       • Pps had to remember a pattern "F"
         • Visualise a dot moving to inner and outer corners
     • Verbal Task
       • Pps asked "is each word a concrete noun? A bird in the hand is not in the bush"
     • Verbal output
       • "Yes, yes, yes"/"no, no, no"
       • Visual output
         • Pointing to "yes"/"no"
     • Findings
       • Interference effect
         • If you have encoded info visually then it is difficult to output verbally, vice versa
           • Both types of info stored seperately

7. Evidence for Phonological Loop

   • Conrad & Hull (1964)
     • Phonological Similarity Effect
       • Poor recall of word lists where items sound similar even when items are presented visually
         • Items encoded according to how they sound
   • Salame & Baddeley (1987)
     • Recall impaired by simultaneous speech
       • Involuntary phonological encoding
   • Baddeley et al. (1975)
     • Serial recall is approximately as many words as you can read out loud in 2 seconds
       • Span is lower for longer words than for short ones, even visually presented
         • Spans are longer for faster speakers
           • Subvocal rehearsal is like vocal rehearsal

8. The Central Executive

   • Based on Norman & Shallice's (1982) Supervisory Attention System (SAS)
   • Schemata
     • Behaviour may be controlled common habits
       • Routine tasks are controlled in a semi-automatic way by behavioural schemata which contain instructions on how to perform a task
         • One or more schemata can be activated to complete multiple tasks
   • SAS
     • Overall controller as a limited capacity attentional component
       • Useful when the individual is confronted by a novel situation for which there is no stored schemata
   • Role
     • Control of behaviour based on action schemas
     • Low level 'contention scheduling' chooses next schema
     • SAS can override the general process of contention schedulling by directly activating or inhibiting schemas
     • SAS failure: frontol love patients' perseveration and utilisation behaviour (Shallice, 1982)
     • Experimental task that seems to load the central executive is random number generation (Baddeley 1966)
     • Random interval generation as a cure for insomnia?

9. Additions to the Working Memory Model

   • WM has problems accounting for semantic/LTS factors
     • Sentence span about 15 words
   • Baddeley (2000, 2003)
     • Interaction with LTM
       • Chunking via Episodic Buffer

Images and files in the map