Shannon Rufus

1. Sensory Memory (SM)#

Information from the environment first enters sensory memory through our five senses.

Key features:

• Acts as a buffer for stimuli received through sensory systems
• Information is unprocessed - it’s a direct copy of the sensory input
• Most information is rapidly lost unless attended to

TIP

Sperling (1960) demonstrated the existence of iconic memory using a partial report procedure - participants could recall any row of letters flashed briefly, proving all letters were briefly stored!

2. Short-Term Memory (STM)#

Information that is attended to enters the short-term memory.

Characteristics:

• Duration: 18-30 seconds without rehearsal (Peterson & Peterson, 1959)
• Capacity: 7±2 items (Miller, 1957) or more recently 4 chunks (Cowan, 2001)
• Coding: Mainly acoustic (Conrad, 1964) - we remember sounds

Maintenance rehearsal:

• Repetition of information keeps it in STM
• Essential for transfer to Long-Term Memory
• Without rehearsal, information is lost through:
  • Decay - Fading of memory trace over time
  • Displacement - New information pushes out old information

Case study: Clive Wearing

• Severe amnesia after herpes virus infection
• STM intact but cannot form new LTM
• Demonstrates separation between STM and LTM stores

3. Long-Term Memory (LTM)#

Permanent storage of information that has been rehearsed and processed.

Characteristics:

• Duration: Potentially unlimited (can last a lifetime)
• Capacity: Potentially unlimited
• Coding: Mainly semantic (meaning-based) though also visual and acoustic

Types of LTM (later added by other researchers):

Sperling (1960) - Sensory Memory#

Aim: To demonstrate the existence of sensory memory

Method:

• Displayed 12 letters in a grid for ~0.5 seconds
• Whole report condition: Recall all letters (avg 4-5)
• Partial report condition: Tone indicated which row to recall (avg 5 items from any row)

Findings:

• In partial report, participants could recall any row
• Proved all letters were briefly stored in sensory memory
• But iconic memory fades too quickly for full report

Conclusion: Sensory memory has large capacity but very short duration

Peterson & Peterson (1959) - Duration of STM#

Aim: To test how long information remains in STM without rehearsal

Method:

• Participants given trigrams (3-letter nonsense syllables)
• Count backwards in 3s to prevent rehearsal
• Recall after 3, 6, 9, 12, 15, or 18 seconds

Findings:

Conclusion: STM duration is about 20 seconds without rehearsal

NOTE

This study used trigrams specifically because they have no meaning - preventing participants from using semantic coding which would improve recall!

Miller (1956) - Capacity of STM#

Aim: To determine the capacity of STM

Method:

• Various experiments with different types of information
• Digit span, letter span, word span

Findings:

• Average capacity = 7±2 items
• This limit applies to various types of information

Chunking:

• Miller found we can remember more by grouping information
• Example: 07911234567 is easier to remember as 07911-234-567
• Chunking increases effective capacity

Bahrick et al (1975) - Duration of LTM#

Aim: To investigate very long-term memory (up to 50 years)

Method:

• Tested 392 American graduates on:
  • Free recall of former classmates
  • Photo recognition of classmates
  • Name recognition from photos

Findings:

Conclusion: LTM can potentially last a lifetime when information is meaningful

1. Brain Damage Evidence#

Case study: HM (Henry Molaison)

• Brain surgery to treat epilepsy (1953)
• Hippocampus removed
• Could form new STM but could not transfer to LTM
• Showed STM and LTM are separate stores

Case study: Clive Wearing

• Herpes simplex virus attacked hippocampus
• Similar to HM - intact STM, severely impaired LTM
• Could not form new memories
• Demonstrates functional separation of stores

2. Experimental Evidence#

Glanzer & Cunitz (1966) - Serial position curve:

1
% Recall
2
   |
3
   │           ╭───── Recency effect (STM)
4
   │          ╱
5
   │         ╱
6
   │        ╱
7
   │       ╱
8
   │      ╱                    Primacy effect (LTM)
9
   │     ╱                    ╲
10
   │    ╱                      ╲
11
   │   ╱                        ╲
12
   │  ╱                          ╲
13
   │ ╱                            ╲──────────
14
   └───────────────────────────────────────── Position in list
15
   1                                   10

• Primacy effect (beginning of list) = Rehearsal → LTM
• Recency effect (end of list) = Still in STM
• Middle items = Neither rehearsed enough nor still in STM

With delayed recall (counting backwards):

• Recency effect disappears (STM displaced)
• Primacy effect remains (already in LTM)

1. Oversimplified#

• Too simple - memory is more complex than just 3 stores
• Doesn’t account for different types of LTM (episodic, semantic, procedural)
• Doesn’t explain why some things are remembered better than others (emotional significance)

2. STM is More Complex#

Working Memory Model (Baddeley & Hitch, 1974):

• STM is not just one store
• Has multiple components:
  • Central executive (coordinates)
  • Phonological loop (acoustic info)
  • Visuospatial sketchpad (visual info)
  • Episodic buffer (integrates info)

3. Rehearsal is Not the Only Way#

Levels of Processing (Craik & Lockhart, 1972):

• Depth of processing matters more than rehearsal
• Semantic processing (understanding meaning) → better recall
• Shallow processing (appearance/sound) → poor recall

Example: Remembering “DOG”

• Shallow: “It has 3 letters” (structural)
• Medium: “It rhymes with LOG” (phonetic)
• Deep: “It’s an animal that barks” (semantic) ✓ Best recall!

4. LTM Influences STM#

Case study: KF

• Motorcycle accident → brain damage
• Poor verbal STM but good visual STM
• LTM influenced STM - contradicts MSM’s one-way flow

4-Mark Evaluation Points#

Strength: Case study of HM

• After surgery to remove hippocampus, HM could not form new long-term memories
• But his short-term memory was intact (could hold conversation)
• This supports the idea that STM and LTM are separate stores
• However, case studies have small samples and may not generalise

Weakness: Working Memory Model

• MSM suggests STM is a single unitary store
• But Baddeley & Hitch’s Working Memory Model shows STM has multiple components
• This makes MSM oversimplified
• Supported by dual-task performance (we can do visual + verbal tasks simultaneously)

Weakness: Rehearsal is not the only factor

• MSM suggests rehearsal is essential for LTM storage
• But Craik & Lockhart found depth of processing is more important
• Semantic processing leads to better recall than simple rehearsal
• This contradicts the MSM

Practice Question#

Outline and evaluate the multi-store model of memory (16 marks)