Memory

Definitions

Related to recognition, language, reasoning, categorisation and planning/problem-solving

Retention of learning and experience

Ways of measuring memory

• recall

• recognition

• memory span

• narrative accounts and subjective experiences, diaries

Memory processes

Encoding

• changing incoming information into form that can be stored

• types of memory code (e.g. acoustic, visual, semantic)

Storage

• retaining information and keeping it accessible ( without deleting or modifying it

• types of long-term memory (e.g. episodic, semantic , procedural etc.)

Retrieval

• being able to remember the experience

• types of retrieval test (e.g. recall vs recognition)

Multi-store Model of Memory

Atkinson & Shiffrin (1968) argued that memory comprises series of stores

Differences betweem STM and LTM

Modal model over simplifies both STM and LTM

• other factors (not just rehearsal) important for storage in LTM

• focuses on structure, not enough attention to processes

Multiple Systems within LTM

Declarative (aka explicit)

• involves conscious recollection ‘knowing that’

Episodic vs Semantic memory (Tulving, 1972)

Record of personal experiences vs general knowledge store

• controversy over the relationship between these two systems and how they interact (Tulving, 2002; Spiers et al, 2001)

Non-declarative (aka implicit)

• doesn’t involve conscious recollection ‘knowing how’

Encoding and Retrieval

Levels of Processing

Proposed by Craik and Lockhart (1972)

• memorability is a function of the level of processing recieved at encoding

Support from Craik and Tulving (1975)

• series of yes/no questions about words followed by recognition test

• questions involved different ‘levels’ of processing

Words encoded shallow level (case and rhyme) remembered less than those encoded deeper level (sentence)

Craik and Tulving (1975)

Strengths:

• first emphasise importance processes in understanding memory (link between encoding and later retrieval)

• stimulated lots of research into memory processes

Limitations

• circulatory of argument - what is deep processing?

• amnesic patients can process semantically (understand conversations) but can’t remember them

• role other variables (elaboration, distinctiveness)

• LOP effects much less with implicit memory tasks (Challis et al, 1996)

Tulving (1979)

Information in memory combination to-be-remembered material and its context

Godden and Baddeley (1975)

Recall much better when environment same at recall test as at encoding

Forgetting - Ebbinghaus (1885/1913)

series of 3 letter words on cards

Influencing factors of forgetting:

Decay

• Passage of time correlated with forgetting

• but, difficult to measure and evaluate

• forget more and forget it faster if engage normal activity than if sleep same amount of time

• suggests inference more important than decay in LTM (Jenkins & Dallenbach, 1924)

Interference

• new information and stored memories compete

• but is interference due to old memory being lost from memory (displaced by new) or is it just harder to locate the old information

Tulving & Psotka (1971)

Lost or just can’t locate?

• Buildings list: hut, cottage, tent, hotel, …

• Geography list: cliff, hill, river, volcano, …

Memory accuracy and errors

Memories are prone to distortion - Bahrick et al, 1996

Flashbulb memories (e.g. Schmolk et al, 2000; Talarico and Rubin, 2003)

• memories are reconstructive

• reconsolidation

Summary

• What is memory & how can it be measured?

• Is memory a unitary construct?

• What evidence exists for a distinction between STM & LTM?

• What distinctions have been suggested within LTM?

• What role do memory processes play in our ability to retain/recall information?

• How & why do we forget? How accurate are our memories?

Muti-Store Model of Memory

Atkinson & Shiffrin (1968) argued that memory comprises series of stores, with each store operating in a single, uniform way

Measuring STM - Digit Span & Corsi Block Task

Short-term memory: One system?

Atkinson & Shiffrin’s model suggests that STM is a single store … BUT

• Shallice & Warrington (1972)

• KF had impaired verbal STM (digit span) but not visuo-spatial (Corsi block)

• Shallice & Warrington (1974)

• KF’s deficit limited to verbal materials (e.g. letter, words & digits) but not meaningful sounds

Working Memory

Baddeley & Hitch (1974); Baddeley (2000)

• Memory is not just a passive store - information is in memory to be used

• Modal model is a linear ‘production line’, whilst working memory can work in parallel and ‘multi-task’

• system(s) within memory involved in the temporary maintenance & manipulation of information

• Example: 26+37?

• Working Memory performance influences number of other cognitive processes

• Language development & learning vocabulary (Gathercole & Baddeley, 1990)

• Maths development (Holmes & Adams, 2006)

• Reading & comprehension (Daneman & Carpenter, 1980; Palmer, 2000)

• Working memory deficits are also linked to learning difficulties such as dyslexia & ASD

VSSP AND PL controlled by CE

PL maintains and manipulates verbal and acoustic material

VSSP maintains and manipulates spatial and visual information

Rehearsal maintains information: decay and forgetting is very rapid without it

Testing the WM

Each component (CE, VSSP, PL , EB) limited capacity

Assumed each component can function relatively independently

Two tasks interfere if they require acccess to a common component of WM and combined demand exceeds component’s capacity

If SHARE: (Task1 + Task2) < Task1 or Task2 separately

If DIFF: (Task1 + Task2) = Task1 or Task2 separately

The Phonological Loop

2 sub-systems: phonological store + articulatory loop

• phonological store: stores traces of acoustic or speech-based material

• articulatory loop: maintains phonological memory traces , recodes visual information by subvocal naming

• recodes visual info through rehearsal

Evidence to support phonological loop:

• Articulatory suppression: prevents rehearsal so information decays and is lost (Baddeley et al, 1975)

• Phonological similarity effect: harder to recall words which sound alike

• Word length effect: span for polysyllabic words shorter than for one syllable words (Baddeley et al, 1975). Why? Rehearsal takes longer

• Irrelevant speech effect: unattended irrelevant speech (foreign language) activates PL, impairing recall

Phonological similarity effect

• Larsen, Baddeley, and Andrade (2000)

• FEE, HE, KNEE, LEE, ME, SHE: phonologically similar

• BAY, HOE, IT, ODD, SHY, UP: phonologically dissimilar

• Serial recall was 25% worse with the phonologically similar list

• Convincing evidence we use the phonological loop on short-term memory tasks

The Visuo-Spatial Sketchpad

• Used to hold and manipulate visual information stored in LTM - visual coding

• Enables us to remember visual and spatial information: maps, directions, geographical orientation...

Useful playing computer games!

Baddeley - listening to American football disrupts driving

Divisions - sepereate components for visual and spatial information

• Logie (1995)

Visual cache: stores info. about colour and shape

Inner scribe: deals with spatial information

Kalauer & Zhao (2004)

• Two main tasks: spatial task (dot memory) and visual task (chinese letters) Secondary tasks: spatial interference (movement) and visual interference (colour)

• More disruption main task when secondary task type (spatial or visual) matches

The Central Executive

• Most important mechanism in the system, but least understood

• Controls action and attention, allocation of resources

• As demand grows, relies on stereotypical or automatic responses

• Evidence: neuropsychological evidence (dysexecutive syndrome) and complex tasks

Baddeley (1996)

• switching of retrieval plans

• timesharing in dual-task studies

• selective attention to certain stimuli while ignoring others

• temporary activation of long-term memory

Baddeley, A. (1996). Exploring the Central Executive. Quarterly Journal of Experimental Psychology, 49A, 5-28.

Smith and Jonides (1999)

• switching attention between tasks

• planning sub-tasks to achieve goal

• selective attention, inhibition

• updating and checking the contents of WM

• coding representations in WM for time and place of appearance

Smith, E. E. & Jonides, J. (1999). Storage and Executive Processes in the Frontal Lobes. Science, 283, 1657-1661.

Episodic Buffer (Baddeley 2000)

• Incorporated into the WM model to deal with the ‘binding problem’

• Integrates information from LTM with information from sub-components of WM

Baddeley, A. (2000). The episodic buffer: a new component of working memory? Trends in Cognitive Science, 4, 417-423.

Summary

• How can STM be measured?

• What evidence exists to suggest that STM is not a unitary store?

• What is the working memory model as outlined by Baddeley & Hitch (1974)?

• How was this model revised by Baddeley (2000)?

• What evidence exists for the WM model & its proposed components?