The Modal Memory Model:
Sensory Memory and Short-Term
(Working) Memory
1
• 1960s Many models of memory
proposed
• Atkinson & Shiffrin (1968)
– Sensory Memory
– Short-term Memory
– Long-term Memory
2
William James
• Primary Memory
• Secondary Memory
3
Atkinson & Shiffrin Model of
Memory (1968)
4
STM Bottleneck
Sensory
Memory
STM
Long
Term
Memory
5
Properties of the Different
Memory Stores
6
Characteristics of the Memory
Stores
7
Research on the A & S Model
• Serial Position Effect
• Recency Effect
• Kintsch & Buschke (1969)
• Behavioral Neuroscience Evidence
8
Serial Position Effect Demo
9
Serial Position Effect Graph
Primacy
Effect
Recency
Effect
10
Rundus (1971)
11
Questions
• How could we test the idea that the last few
items are in STS?
• How can we test that the primacy effect
represents LTS?
12
Eliminating the Recency Effect
13
Other Evidence: Kintsch & Bushchke
(1969)
14
Behavioral Neuroscience Evidence for
the STM-LTM Distinction
• H.M.
- Epileptic
- Temporal Lobes / Hippocampus
- STM ---> LTM disrupted
• K.F.
- Damage to Left Cerebral Cortex
- LTM Normal
- STM capacity severely limited
15
Behavioral Neuroscience Evidence for
the STM-LTM Distinction
The dog bit the man and the man died.
vs.
The man the dog bit died.
16
Evidence Against A & S
• More recent research challenges the strict
coding distinction
• Recency Effect challenged
• Neuroscience evidence
17
Atkinson & Shiffrin Model of
Memory (1968)
18
The Sensory Store
19
Lightning
20
Lightning Demo
21
Lightning Questions
22
Sensory Memory
• Sensory memory or sensory register
• Visual, auditory, touch, taste, smell
• Relatively raw, unprocessed form
23
Why Do We Need Sensory Memory?
• Stimuli change
• Maintain for selection and further
processing
• Integrate fragments of a stimuli into a single
unitary perception
24
Classic Studies
• Sperling (1960)
• Averbach & Sperling (1961)
25
A Tachistoscope
26
27
*
28
J
Z G B
S
X P L
R M Q F
29
30
31
*
32
Y Q C H
N D R J
V B K S
33
34
Schematic of Typical Sperling Exp
35
Number of Letters Recalled as a Function of
Technique & Delay
36
Iconic Memory
1. Location
2. Usefulness
3. Saccades
4. Nature of the code
37
1
K 5 L
H
J 3 B
7
D 8 T
38
Demo 4.1: Examples of Sensory
Memory
39
Demo 4.2 Unitary Perception
from Fragments
40
Auditory Sensory Memory
• Neisser (1967) - Echoic memory and the echo
• Darwin, Turvey, & Crowder (1972)
• Differences from iconic memory
• Crowder (1982)
41
An Echoic Memory Study
42
Darwin, Turvey, & Crowder
43
Discriminating Between Two
Sounds (Crowder, 1982) Graph
44
Short-Term Memory
45
Short-Term Memory
• Nature of Forgetting
• Duration
• Nature of Code
• Capacity
46
Short Term Memory
• Brown/Peterson & Peterson (1959)
• Trigram task
KHR
0 – 18
seconds
Delay / Distractor
(947, 946, 945
. . . 939)
Recall Trigram
47
Trigrams
K X J
P L G
S Y T
H Z R
48
Brown-Peterson Results
49
STM--Nature of the code
• Conrad (1964)
• Visual display of letters
• Phonological confusions: (‘D’ for ‘E’
but not ‘F’ for ‘E’)
• Wickelgren (1965)
50
Wickelgren (1965)
K Z L F
distractor tasks
(copy down 4 ne w letters)
C B G D
X M I
W
recall original 4 letters
51
STM Capacity Limited
Sensory
Memory
STM
Long
Term
Memory
52
Capacity of STM
• Limited Capacity (7 + 2)
• Digit Span Task
• Difficulties
53
Capacity of STM (cont.)
• Chunking
• Recoding:(1 4 9 2 ----> ‘1492’ Columbus)
• Chase & Ericsson (1982)
54
SF DIGIT SPAN DEMO
55
SF Digit Span Experiment
•
•
Initial Session (8 digits):
Digit Series:
SF’s Recall:
Later Session (11 digits):
Digit Series:
SF’s Recall:
SF’s Report:
•
1, 0, 5, 3, 1, 8, 7, 4
105
31874
90756629867
907
566
29867
9:07 a 2-mile time
Still Later Sessions (22 digits):
Digit Series: 4131778406034948709462
SF’s Recall: 413.1 / 77.84 / 0603
494 / 870 / 946.2
SF’s Report: 4:13.1 mile time
06:03 mile time
9:46.2 2-mile time
56
Revisions to the STM Idea
57
Brown & Peterson Revisited
• Decay vs. Interference
• Waugh & Norman (1965) - Probe digit task
• Varying the type of distractor task and
stimulus material
• Keppel & Underwood (1962)
• PI = Proactive Interference
• Wickens et. al. - Release from PI
58
16 digits
----->
probe digit
519635142867394
9837571493862752
59
Effect of Presentation Rate vs. Number of Interfering
Items on Recall (Waugh & Norman, 1965)
60
Wickens, Born, & Allen (1963)
Trial 1 – ‘HJX’
Trial 2 – ‘RLB’
Trial 3 – ‘ZNF’
Control
‘GST’
Experimental
‘493’
61
Release from Proactive Interference
62
Release from PI (Evidence for Semantic Codes)
63
Release from PI as a Function of Semantic
Similarity (Based on Wickens, et al., 1976)
64
Working Memory
• Revision of STM
• 3 part system
• Baddeley
• Dual task paradigm
65
Baddeley Working Memory Model
66
Reasoning Task with Letter Recall
AB
‘A’ precedes ‘B’?
T or F
‘B’ is preceded by ‘A’ .
T or F
‘B’ does not precede ‘A’.
T or F
67
Reasoning Speed and Letter Recall
Experiment 1:
0, 1, 2 items preloaded
reasoning task
letter recall
Experiment 2:
0 or 6 items
reasoning task
letter recall
68
Reasoning Times & Letter Recall Results
69
Bradimonte Et al. (1992)
70
Brandimonte (1992)
Condition 1
1. Study 6 pictures
Condition 2
1. Study 6 pictures while
saying “la, la, la . . .”
2. Create mental image,
subtract a specific part,
and name it.
2. Create mental image,
subtract a specific part
and name it.
?
3. Number of correct
items: 2.7
Fish
3. Number of correct
items: 3.8
71
Pronunciation Time & Memory Span
72
Memory Span and Pronunciation Rate
73
Capacity of STM
 Difficult to Estimate
 Different meanings (storage capacity vs.
processing cap acity)
 Digit Span Task
 Miller – “The Magical Numb er Seven,
Plus or Minus Two . . .”
7±2
74