Working Memory Components Evidence Different codes Dual-task paradigm Dissociations.

Working Memory

Components

EvidenceDifferent codesDual-task paradigmDissociations

Primary Memory Secondary Memory

Short-term Store Long-term Store

Short-term Memory Long-term Memory

Immediate Memory(Miller, 1956)

*Working Memory

Info InfoShort-termMemory

Long-term

Memory

Info InfoWorkingMemory

Long-term

Memory

Baddeley & Hitch (1974)

WorkingMemory

Long-termMemory

Baddeley & Hitch (1974)

Baddeley & Hitch (1974) -- Working memory

Baddeley & Hitch (1974) -- Working memory

Long-termMemory

Baddeley & Hitch (1974) -- Working memory

Long-termMemory

CentralExecutive

Baddeley & Hitch (1974) -- Working memory

Long-termMemory

CentralExecutive

PhonologicalLoop

Visuo-spatialSketchpad

Baddeley & Hitch (1974) -- Working memory

Long-termMemory

CentralExecutive

*PhonologicalLoop

Visuo-spatialSketchpad

*also called Articulatory-

Rehearsal Loop

Baddeley & Hitch (1974) -- Working memory

Long-termMemory

CentralExecutive

*PhonologicalLoop

Visuo-spatialSketchpad

*also called Articulatory-

Rehearsal Loop

Slavesystem

Slavesystem

Baddeley & Hitch (1974) -- Working memory

Long-termMemory

CentralExecutive

*PhonologicalLoop

Visuo-spatialSketchpad

Slavesystem

Slavesystem

Info Response

Working Memory

Why divide STM into different components?

Evidence Different codes

Phonological effects with visual stimuliRelease from PIMental rotation of visual stimuli

Different codesPhonological effects with visual stimuli

Conrad (1964)immediate recall of visual stimuli (letters)noticed phonological errors in recall protocols

(ē confusions, ĕ confusions)

Different codesPhonological effects with visual stimuli

Conrad (1964)immediate recall of visual stimuli (letters)noticed phonological errors in recall protocols

(ē confusions, ĕ confusions)

Wickelgren (1965) visual presentation of letters;

copy letters during retention interval; recall first letters presented phonological interference effects

Different codesPhonological effects with visual stimuli

Verbal code/phonological code

Verbal stimuli get coded phonologically

Different codes Release from PI

Release based on meaning of stimuli

Semantic code

(note: reflects an interaction with long- term memory)

Baddeley & Hitch (1974) -- Working memory

Long-termMemory

CentralExecutive

PhonologicalLoop

Visual-spatialSketchpad

Access meaning stored in long-term memory

Different codesStrongly visual stimuli (shapes without meaning)

Perform mental rotation taskmanipulate degree of rotation

See overhead transparencies

RT(s)

Reaction Time for “Same” Judgmentsas a Function of Angle of Rotation

Angle of Rotation (in degrees)

0 45 90 135 180

1

2

3

5

4

Different codesStrongly visual stimuli (shapes without meaning)

Perform mental rotation taskmanipulate degree of rotation

Greater degree of rotation, the longer the RT To do the task, we need a visual code

Working Memory

Why divide STM into different components?

Evidence

Dual-task paradigm

Dual-task paradigm

Do two things at once

One task – measure accuracy or RT

Second task – manipulate task or task difficulty

Dual task paradigm Baddeley & Hitch (1974)

Reasoning task – measured RTsee ABnext, get a sentence: A precedes B(respond “yes” or “no”)

Second task – do nothing (control) vs. hold a random 6-digit sequence in mind

Reas oning

Reasoning Time as a Function of Second Task

Sentence Type

1

2

3

5

4

6

Ti

me

(in seconds)

Active

Affirmative

Random digitsControl (no 2nd task)

Longer RT for dual-task condition

Only so much capacity to do mental work2nd task hurts performance on 1st task

Reasoning task – measured RT also manipulated sentence type:Affrimative vs. NegativeActive vs. Passive

AffirmativeActive: A precedes BPassive: B is preceded by A

NegativeActive: B does not precede APassive: A is not preceded by B

Reas oning

Reasoning Time as a Function of Sentence Type and Second Task

Sentence Type

1

2

3

5

4

6

Ti

me

(in seconds)

Active Passive Active Passive

Affirmative Negative

Random digitsControl (no 2nd task)

As 1st task gets harder, two things happen 1) RT increases, in general

2) Second task, if difficult, particularly hurts performance in the harder conditions of the 1st task

(e.g., 1st task, negative passive condition 2nd task, hold random digit sequence in mind)

Working Memory

Why divide STM into different components?

Evidence

Dissociations

Working Memory

DissociationsUse dual task paradigm

Reasoning 2nd task should interfere more with 1st task if

the same slave system is used for both tasks (less interference if different slave systems areused)

Working MemoryDissociations (Logie, Zucco, & Baddeley, 1990)

Measure visual spansee grid of blocksgrid change (one block different)point to change

Second taskMental additionImaging

Working Memory

Dissociations Measure visual span – visuo-spatial sketchpad

Second task Mental addition – phonological loop Imaging – visuo-spatial sketchpad

Working Memory

Dissociations Measure visual span – visuo-spatial sketchpad

Second task Mental addition – phonological loop Imaging – visuo-spatial sketchpad

Prediction – Should be a greater cost in visual span performance for the imaging task

Cost in Performance as a Function of Task Combination

Task Combination

0

70

Addition Imagery

Visual span task

Per

cent

Dro

p f

rom

Bas

elin

e

Working Memory

Dissociations Measure visual span – visuo-spatial sketchpad

Second task Mental addition – phonological loop Imaging – visuo-spatial sketchpad

Prediction upheld: Bigger cost in imaging condition

Working Memory

Dissociations Measure letter span – phonological loop

see (or hear) letters; hold letters in mind;then recall letters

Second task Mental addition – phonological loop Imaging – visuo-spatial sketchpad

Prediction – Should be a greater cost in letter span performance for the mental addition task

Cost in Performance as a Function of Task Combination

Task Combination

0

70

Addition Imagery Addition Imagery

Visual span task Letter span task

Per

cent

Dro

p f

rom

Bas

elin

e

Working Memory

Dissociations Measure letter span – phonological loop

Second task Mental addition – phonological loop Imaging – visuo-spatial sketchpad

Prediction upheld: Bigger cost in mental addition condition

Working Memory

Components

EvidenceDifferent codesDual-task paradigmDissociations