Memory

• Memory


  3 processes:

  1. Acquiring information

  2. Storing information

  3. Retrieving information

  
  

  capacity - amount of info that can be stored

  duration - length of time it can be stored for

  encoding - way info is stored.
• Sensory memory


  Sperling 1960

  C - very large

  D - milliseconds

  E - mainly visual (echoic and haptic)
• STM


  C - Miller, 7+/-2 or 5-9 chunks

  D - 18-30 seconds, Peterson and Peterson

  E - Acoustic, Baddeley
• LTM


  C - unlimited

  D - life time, Bahrick

  E - semantic, Baddeley
• Sensory memory


  Holds info in relatively unprocessed form for fractions of a second.

  Allows successive eye fixation to become intergrated, a continuous image.

  iconic store - visual, short period of time

  echoic store - auditory, milliseconds

  haptic store - touch and muscle tension
• Sperling 1960 - sensory memory

  Studied the iconic store; flash stimuli on a screen for short instances.

  - p's asked to remember as many letters as possible from 12 symbol grid displayed for 50ms.

  - could only remember around 4.

  capacity: used the 'partial report technique', used 3 tones and when grid disappeared a tone would indicate which row of 4 letters. Recall was just over 3 of 4 symbols

  - iconic store can retain 76% of data.

  Duration: If there was a delay between gird and tone more info was lost, 33% after 1 sec delay.
• Capacity of STM - Miller


  7+/- (5-9)

  research using increasing number digit spans, p's would recite sequence until couldn't remember, in pairs.
• Stratergies to increase STM

  - chunking (5-9): combining letters etc e.g telephone

  - rythmic grouping

  - utilising LTM
• Duration of STM - Peterson and peterson


  A = test duration

  P = show constant triangle, asked to count backwards in threes from a number to stop rehearsing. after intervals of 3, 6 or 18 seconds asked to recall. repeated.

  F = recall 80% after 3s, fewer recalled as intervals lengthened, 18 second less than 10% recalled

  C = STM lasts 18-30 seconds

  E = artificial and possible interference of previous trigrams, only one type of stimuli, lacked mundane realism.

  - experimental method, cause and effect (IV & DV), scientific, repeatable, standardised procedure
• Encoding of STM - Baddeley (1966)


  A = effects of acoustic and semantic encoding

  P = 4 groups, list of 5 words

  1. acoustically similar, 2. acoustically dissimilar, 3. semantically similar and 4. semantically dissimilar.

  - Immediately asked to recall

  F = acoustically similar harder, 55% accuracy.

  C = STM relies on sound of words more than meaning so acoustically encoded.

  E = experimental method shows causal link

  - simplistic (Brandimont; may use other forms e.g visual)
• Support - Conrad


  P = p's shown sequence of 6 consonants, projected quickly on screen. Either acoustically similar or not. Wrote down in appearance.

  F = most errors were substitutes of similar sounding letters. more difficult to recall acoustically similar words in order.

  C = STM acoustically stored
• Capacity of LTM


  Huge
• Duration of LTM - Bahrick


  A = Establish existence of LTM and difference between recognition and recall.

  P = 392 graduates from American high school over 50 years.

  recognition group: match list of names to photos

  recall group: no list of names

  F = Recognition: 90% correct 14 years after, 60% after 47.

  Recall: 60% after 7 years, less than 20% after 47.

  C = LTM better when measured through recognition, can remember some info for a lifetime.

  E = Was it a decline in age or memory?

  - meaningful stimulus
• Encoding of LTM - Baddeley


  Same study as encoding of STM

  F = recall worse, 55% for semantically similar.

  C = LTM semantically encoded, STM is acoustic.

  E = causal link

  - Only tests words e.g not visual.
• Multi store model of memory - Atkinson and Shiffrin


  Suggests memory is a linear flow of information, passes through stages.

  
• MSM sensory memory


  1. sensory memory

  - external stimuli enters uncoded.

  - registered for milliseconds

  ATTENTION vital to be passed to STM
• MSM STM


  2. STM

  - store 7+/-2 items in acoustic code for 18-30 seconds.

  - must be REHEARSED

  - memory lost due to displacement of new info or decay.
• MSM LTM


  3. LTM

  - unlimited amount and can last a lifetime

  - semantically encoded

  - forgotten due to decay, interference of new info or retrieval failure.
• MSM evaluation - HM (case study)

  Suggested we have separate memory stores as his LTM was fine.

  - Brain surgery to remove hippocampus and alleviate epilespy.

  - Unable to transfer STM to LTM, no knowledge of current affairs. But could remember people from long ago.

  Supports MSM
• MSM evaluation - Clive Wearing (case study)


  Impaired STM, duration of 2 seconds.

  - unable to rehearse new info or transfer LTM

  - difficulty retrieving LTM but could remember procedural memories.

  Supports as demonstrates STM & LTM are different stores as one impaired more than other.
• MSM Evaluation (case study)

  - Simplistic

  - multiple stores within e.g Clive W only some of his LTM was impaired.
• MSM evaluation - Glanzer and Cunitz (support)


  Lab experiment (1966)

  A = evidence for separate STM and LTM stores.

  P = list of words and recall in any order. Immediate recall group & delayed recall group (count backwards for 30s).

  F: Immediate - remembered 1st and last words best (primary effect and recency effect).

  Delayed - remembered words at beginning best.

  C = Both remembered words at beginning as stored in LTM, words at end still in STM.

  - when recall delayed by task unable to rehearse, words not maintained in STM or transferred to LTM.

  Support idea of separate stores: delayed recall caused recency effect to disappear but no affect on LTM and left primacy effect.

  E = artificial, lab, tasks lack mundane realism, low val.

  - scientific and reliable
• Criticism of MSM


  - Simplistic, evidence CW suggest different stores within STM and LTM.

  - Only focuses on process of remembering and amount that is produced not what is processed. E.g something was just remember.

  - Suggests rehearsal is only way to transfer STM to LTM. Not true.
• Working memory model - Baddeley and Hitch

  - Just STM

  - Thought MSM was too simplistic.

  - actively processing info

  - identifies we could multi task

  
• central executive


  Most important - supervisory system to slave systems

  - controls info flow, decisions and attention

  Flexible = can process an form of info

  capacity = limited
• Phonological loop


  Sound based : stores limited number of speech based sounds for brief time.

  phonological store (inner ear) - acoustically coded items to be stored, limited capacity, brief time.

  Articulatory control process (inner voice) - silent speech, internal speech, allows repetition of items to be stored.
• Visuospatial scratch pad


  stores visual and spatial info 'inner eye'

  - mental image e.g navigation

  - limited capacity
• Episodic buffer


  retrieve material from LTM to meet requirements of working memory, combine and integrate new info.

  - limited capacity.
• Evaluation of WMM (support)

  - research agrees STM is made up of components.

  - explains more than MSM

  - real life application: phonological loop - reading, central executive - problem solving, VSSP - navigation.

  - supported by dual task studies Baddeley and Hitch
• Evaluation of WMM (against)

  - little evidence of how the central executive works and its capacity, vague and unstable nature.

  - only involves STM

  - Liebermann; blind people have good spatial awareness despite never having visual info. argue VSSP should be separated to visual and spatial info.

  - not unitary, HM not all stores within impacted
• Types of LTM - Endel Tulving (MSM of LTM too simplistic)


  1. episodic

  2. semantic

  3. procedural
• Episodic memory


  Ability to recall events from our lives.

  - time stamped; remember when they happened

  - include several elements e.g people that make an event.

  - conscious; have to make a conscious effort to recall, quick but aware.
• Semantic memory


  Knowledge of the world e.g meaning of words.

  - not time stamped, don't remember when first learnt

  - less personal, large collection constantly added to
• Procedural memory


  Actions and skills.

  - recall unconsciously

  - often skills that are hard to explain

  - not time stamped
• Evaluation (support) - clinical evidence

  Case studies of HM and Clive Wearing

  - episodic memory in both severely impaired but semantic relatively fine. E.g HM couldn't recall stroking a dog but knew what a dog was.

  - procedural memories intact e.g CW played piano.

  Supports Tulving, clear evidence of different stores in LTM
• Evaluation (support) - neuroimaging evidence

  Evidence of brain scan studies that types are stored in different areas.

  Tulving - Performed memory tasks and scanned with PET scanner.

  - episodic: right pre frontal cortex

  - semantic: left pre frontal cortex

  Supports view of different stores.
• Evaluation (support) - real life application

  - can target types of memory in treatment.

  Belleville et al

  episodic memories could be improved in older people with mild cognitive impairment. Trained p's performed better in test after training than control group.

  - episodic memory most often affected, highlights benefits of distinguishing types.
• Evaluation (against) - clinical evidence


  lack of control variables
• Evaluation (against)


  3 or 2 types of LTM?

  - argue episodic and semantic stored together = declarative memory, consciously recalled.
• Forgetting - interference


  Two pieces of info cause distortion. forgetting LTM more likely as we cant access it.