Working Memory & Long-Term Memory

• Working Memory is the topic
• Exam 1 Location: Most people will be in 1324 EH, others are in Ross 1100

How to Use the Review Sheet

• Tell a story about each term
• For each term:
  • Define it
  • Know evidence for/against it
  • Explain how it relates to other terms

Lecture 9 Outline: Working Memory

• Issues to consider: Role of WM in thinking and the relationship between WM & LTM
• Centrality of WM in thinking:
  • Raven's progressive matrices
  • Mental arithmetic
  • Individual differences (Raven's, age, and working memory)
• Dissociating WM and long-term memory:
  • Neurobiological evidence (amnesia, WM disorder, double dissociations)
  • Behavioral evidence (serial position, primacy, recency, double dissociations)
• Raven Progressive Matrices are a test of cognitive ability

Lecture Outline

• Tripartite theory of WM includes:
  • Phonological loop
  • Visuospatial sketchpad
  • Central executive

Mental Arithmetic

• Example problem: 836 + 429 involves using long-term memory for strategies and facts, and working memory to store numbers during the process.
• Working memory is a key aspect of mental arithmetic

Working Memory: "Reading Span"

• Subjects read a series of sentences before recalling the final word of each sentence in order
• WM span is the number of final words which can be correctly remembered

Raven’s Progressive Matrices

• Raven's scores are worse for older people
• Raven's gets worse because WM gets worse
• Older people with the same working memory score perform similarly to younger people with that score

Logical Syllogisms

• Example:
  • All men are mortal
  • Professor Lee is a man
  • Therefore, Professor Lee is mortal

Interference Effects in Reasoning

• Syllogisms and random number generation create interference
• Working memory task interference with reasoning

Reading Comprehension

• People with high working memory spans exhibit better comprehension of text with a correlation of 0.72

Anterograde Amnesia

• Anterograde amnesia is the inability to remember events occurring after brain injury, as seen in patient H.M.

Neurobiological Evidence

• Patient H.M had anterograde amnesia, impaired long-term memory, but normal working memory
• Patient K.F had normal long-term memory but impaired working memory

Double Dissociations

• General Logic: If System A is responsible for Task 1 (e.g. LTM), and System B handles Task 2 (e.g. WM)
• Lesion Studies:
  • If a lesion in area A affects System A's performance on Task 1 but not Task 2.
  • If a lesion in area B affects System B's performance on Task 2 but not Task 1

Serial Position Curve

• Primacy effect: enhanced recall of items at the beginning of a list
• Recency effect: enhanced recall of items at the end of the list
• Proportion recalled affected by list length for 20 and 30 word lists

Serial Position Curve

• Rate of Presentation: Fast, Medium, Slow effects recall
• A slower rate of presentation enhances recall for the first items

Lecture Outline

• Tripartite theory of WM
  • Phonological loop
  • Visuospatial sketchpad
  • Central executive

Working Memory Model (Baddeley)

• Consists of the central executive which supervises both the visuospatial buffer and phonological buffer

The Phonological Loop

• Phonological short-term store holds auditory information, which can be maintained through subvocal rehearsal
• Non-auditory inputs and auditory speech inputs feed it

Phonological Coding

• Span test: Listen to the list of letters before writing them down in order
• Acoustic Similarity (Conrad) found more confusion with letters that sounded alike

Acoustic Similarity (Conrad)

• Confusions occur with words that sound alike: mad, cat, man, map, cat
• This is not present for words with similar meaning or appearance

Articulatory Suppression

• Saying "the" repeatedly reduces the formation of phonological code
• Articulatory suppression reduces acoustic confusion effects and increases errors

Articulatory Duration Effect

• Individuals can generally remember as many words as they can say in 2 seconds
• Short words are easier to remember than long words and this effect disappears with articulatory suppression

Speed of Speech Effect

• Memory span is better for words are pronounced quickly
• Memory span is better for people who speak quickly
• Chinese span is better than English span, which is better than Welsh span

The Phonological Loop

• See previous notes above

Chunking

• The strategy of grouping information to increase memory capacity
• Chunking example: FBI CIA FDR JFK

PET Subtraction: Rehearsal

• Mental processes in 2-back task
• Mental processes in search task
• Rehearsal process in 2-back
• PET activation in 2-back task & search task
• Brain areas related to rehearsal

Brain Areas Related to Phonological WM

• Top of head shown on a left to right view
• Areas active in 2-back aren't in search

FMRI Meta-Analysis

• Study Verbal N-Back Tasks
• Brain is measured in mm

The Visuospatial Buffer

• Component of WM devoted to visual imagery and spatial processing
• Receives information from vision and long-term memory

The Visuospatial Buffer

• Defined as the component of WM that is devoted to visual imagery and spatial processing
• Information can enter the buffer directly from visual perception or from long-term memory before being treated like a percept
• A percept can be, scanned, rotated, enlarged etc.

Double Dissociation: Visuospatial Sketchpad vs. Phonological Loop

• Behavioral evidence from Brooks (1968)
• Biological evidence from Jonides et al. (1993)
• Two forms tested in Brooks (1968):
• visuospatial task
• phonological control task
• Dependent variable= time to finish responding

Verbal/Visuospatial Task

• Verbal tasks such as "A bird in the hand is in the bush"
• Visuospatial Response Point to the answer for "A bird in the hand is in the bush"

Brooks Task Interference Effects

• Completion Time is measured against block.
• Spatial WM Task tested by Memory and Control condition

PEW Subtraction: Spatial Working Memory

• Measures mental processes in Memory and Control Condition
• Measures Visuospatial sketch-pad processes
• Measures PET activity in memory and control
• Measures Brain areas related to visuospatial sketch-pad
• Memory tested against control

Areas Active in Memory But Not in Control

• Measured by: Left/ Right, Prefrontal, Premotor, and Occipital

A Double Dissociation

• Studies phonological and visuospatial WM
• Prefrontal and Occipital WM are observed in studies

The Central Executive

• Brains super vision module
• Is responsible for planning and coordination
• Is responsible for monitoring

Frontal Lobe Syndrome

• Consists of distractibility, difficulty in concentrating
• Causes problem in organization and planning
• Causes perseveration: fail to stop inappropriate behavior

Working Memory Model

• Consists of Visuospatial Buffer, Central Executive and Phonological Buffer

Sensory (sensorimotor) Recruitment Hypothesis

• Connection between multiple brain regions
• Explains that different regions are responsible for different senses

Distributed WM Representations

• Consists of Monkey electrophysiology and Human imaging which include complex areas