Block 3 - Memory

• Stages of memory


  1. Encoding

  2. Storage

  3. Retrieval
• Encoding


  Converting info into a useable form
• Storage


  Holding info in memory
• Retrieval


  Taking memories out of storage

  
• Patient HM


  • Had severe epilepsy so had bilateral hippocampus removal
  • Severe changes in memory after surgery
  • Could retrieve long-term memories (LTM) and still had semantic memories
  • Could not lay down new LTMs

  
• Anterograde amnesia


  Loss of ability to create new LTMs after the event that caused the amnesia
• Retrograde amnesia


  Loss of memory access before the event that caused amnesia
• HM could not remember 11 months before surgery but could remember what happened from childhood to early adulthood
• HM's amnesia suggests a distinction between encoding and retrieval (could retrieve old memories but not encode new ones)

  
• Short term memory (STM) and procedural memory was still intact for HM
• Short term memory (STM)

  • Capacity: Miller's magic number (7 ± 2)
  • Duration: 15 - 30 seconds, longer with rehearsal
  • Affected by concurrent task during storage
• Long term memory (LTM)


  • Capacity: Standing magazine pictures (Ppts were able to recognise magazine images seen with 83% accuracy)
  • Duration: Permanent
  • Not affected by concurrent task during storage
• Sperling (1960) found a memory limit, not a perceptual limit, where people tend to remember 4.5/12 letters: span of immediate memory

  
• Sperling's full report experiment

  1. Shown array of letters in 50ms

  2. Asked to report every letter
• Sperling's partial report experiment

  1. Shown array of letters in 50ms

  2. Given an order cue to report a subset of items

  3. People could remember all 4 letters

  4. Partial report superiority - Memory capacity was higher with a cue
• Sensory memory


  • Holds items & positions (e.g. colour, shape, size, brightness)
  • Pre-categorical - only holds physical aspects of a stimuli
  • Large capacity, very short time span

  
• Sensory memory can be extended in duration using a dark field
• Kahneman's integration hypothesis


  Two events combine as one 'psychological moment'
• Visual masks


  Interrupt processing by either replacing or curtailing stimuli
• Visual masks reinforce that sensory processing is a visual phenomenon
• Sperling found that a visual mask still influences visual processing after info from two eyes has been combined, and this cuts short the duration of sensory memory

  
• Summary of sensory memory

  • Partial report superiority
  • High capacity, short time period
  • Pre categorical - only holds physical aspects of a stimuli
  • More than visual persistence

  
• Broadbent's three systems (stores)

  • S-system (sensory memory, unlimited capacity)
  • P-system (limited capacity, items processed and fed back/rehearsed)
  • Long term memory

  
• Broadbent's dichotic listening task

  1. Simultaneously sent one message to right ear and different message to left ear

  2. Physical characteristics of messages enter an unlimited capacity sensory buffer

  3. One of the inputs is selected based on e.g. pitch or loudness for further processing (selective filter - attention/ what you are listening to)

  4. If you pay attention, selected info goes into short term memory and processed semantically

  5. Info that is unattended to is blocked

  
• The cocktail party effect is an example of getting information from other things even if you attend to something else
• Atkinson-Shiffrin Modal model


  1. The process of rehearsal keeps info in STM

  2. Same memory constraints apply to visual, auditory and haptic stimuli

  3. Control processes (encoding, rehearsing, retrieval) regulate the flow of information between stores
• Unitary models unify mechanisms of STM and LTM; see memory as a continuum rather than there being differences
• Distinction between STM and LTM

  • STM seen as fixed number of 'slots'
  • When STM is full, new item must displace old one
  • Displaced items can be sent to LTM, enter rehearsal loop, or be lost/forgotten
  • For STM, items are lost through the process of decay
  • For LTM, items are lost by interference
• Serial order task


  1. Items in the middle of a list are not remembered as well

  2. Shows probability of recall correctly

  3. Primacy effect - superior memory for first items

  4. Recency effect - superior memory for last items

  
• Murdock (1962) found a recency effect regardless of list length or presentation rate

  
• Glanzer & Cunittz (1966)


  1. Recency effect: 0 = immediate recall after list ended, 10 - 10 seconds of counting backwards after list ended, superior performance for immediate recall

  2. Primacy effect: 3, 6 or 9 seconds per item, better performance for 9 seconds per item because there is more time to establish the list in LTM

  
• Problems with the distinction between STM and LTM: Bjork & Whitten (1974) found primacy effects even when rehearsal time was manipulated, Baddeley & Hitch (1977) found recency effects for LTM, Brown (1958) found performance was worse when rehearsal was blocked, Peterson & Peterson (1959) found decay even when items were below immediate memory span
• Kepel & Underwood (1962) replicated the Brown-Peterson task and found performance was not affected over number of trials, and first trial had same correct responses regardless of 3 or 18 seconds
• Proactive interference


  Negative influence of old material on new material
• Waugh & Norman (1965) found that interference plays an important role in how we forget info in STM, challenging Broadbent's ideas