Lecture 5- Intro to Memory
What is memory?
-Memory is essential to all our lives. -Memory is involved in processing vast amounts of information =>e.g. images, sounds or meaning. -Memories are not perfect; they change over time -Memories not stored like books in a library It can be thought of in general terms as the use of past experience to affect or influence current behaviour. Lose detail, become distorted, Shaped by beliefs and expectations, Modified to include new information, not stored like books in a library, Rather a jigsaw puzzle with different elements stored in different parts of the brain. Memorable event - 10 years ago birthday - what you had for supper on Thursday eve. - which is easiest to remember? Advertising and marketing - meaningful and distinctive imagine life without forgetting
Empirical evidence for VSSP
1. Active Interference (like Articulatory Suppression) - imagery mnemonics is disrupted by spatial tracking (Baddeley and Lieberman, 1980). 2. Passive Interference (like Irrelevant Speech Effect) - imagery mnemonics is disrupted by irrelevant visual stimuli (Logie, 1986). 3. Visual Similarity Effect (like Phonological Similarity Effect) - children poorer and remembering visually similar stimuli (Hitch et al, 1986). 4. Visual Complexity Effect (like Word Length Effect) - more complex spatial locations harder to recall. (Snodgrass et al 1980) VSSP has separate mechanisms to process information; visual format and spatial format -No interference between visual and spatial tasks when performed together (Quinn and McConnell 1986) -'What' (ventral stream) and 'where' (dorsal stream) information encoded separately in visual system -Results from brain-imaging show that working memory tasks =>with visual objects activate mostly left hemisphere =>with spatial information activate right hemisphere (PET - Smith and Jonides 1997)
Empirical evidence for PL
1. Dual task paradigms Farmer, Berman & Fletcher (1986) Look at a picture (visuo-sketch pad) and listen to music (Phonological loop) =>Successful, can attend to both; from different components Read a passage (phonological loop) and articulate 'the' 'the' 'the' (articulatory phonological loop) =>Unsuccessful, cannot attend to both; from the same component. 2. Phonological similarity effect Larsen et al. (2000) Serial recall of visually presented words is worse when the words are phonologically similar than phonologically dissimilar. =>Phonologically similar; FEE, HE, KNEE, LEE, ME, SHE =>Phonologically dissimilar; BAY, HOE, IT, ODD, SHY, UP Recall was 25% worse in phonologically similar condition; -Speech-based rehearsal processes within phonological loop. -Implies acoustic trace and encoding 3. The word-length effect Baddeley, Thomson and Buchanan (1975) Participants study words of various lengths (from 1 to 5 syllables) Memory performance is inversely proportional to word length -The longer the word the worse the memory performance 4. Irrelevant speech/Articulatory suppression Salamé and Baddeley (1982) Serial digit recall declined when irrelevant speech sounds played concurrently to visually presented digits. =>The effect is limited to speech sounds =>Do not need to understand what is being said (Colle & Welsh, 1976), even hearing irrelevant speech in another language still disrupts processing Ppts recite irrelevant speech continuously 'the, the, the, the. . . over and over again' -Performance declines It is argued that suppression compromises the ability to recode the visually presented digits into a phonological code; Articulatory suppression.
Empirical evidence for CE
Baddeley & Wilson (1988) Patients with damage to the frontal lobes should therefore behave as if they lack a control system; Dysexecutive syndrome Impairment affecting executive function characterised by; -disturbed attention, increased distractibility, a difficultly in grasping the whole of a complicated state of affairs...well able to work along old routines...(but)...cannot learn to master new types of task, in new situations...[the patient is]...at a loss' (Rylander, 1939, p.20). -In other words, patients suffering from frontal lobe syndrome lack flexibility and the ability to control their processing resources, functions attributed to the central executive. The patients who display utilization behavior tend to reach out and begin to automatically use objects in the visual field of their environment. This may not seem incorrect but the difference in action to a person without UB is that the "object-appropriate" action taken is performed at the inappropriate time. For example, a patient in a doctor's office sees a toothbrush and will involuntarily start brushing his teeth. This demonstrates the appropriate action (brushing) at the inappropriate time (office).
Empirical evidence for CE- Neurophysiological evidence
Baddeley (1996) studied patients with Alzheimer's disease, (reduced CE functioning), -marked reduction in performance when given two simultaneous tasks; evidence for the CE function of distributing attention between two tasks. D'Esposito et al. (1995) dorsolateral prefrontal cortex is activated during dual-task conditions (fMRI). -The Stroop effect requires CE functions of attention and inhibition. =>PET studies have found an area close to the dorsolateral prefrontal cortex (the anterior cingulate gyrus) is active when the colour word conflicts with the print colour.
Working Memory Model
Baddeley and Hitch (1974) questioned the concept of short-term store as proposed by Atkinson and Shiffrin's theory -In a series of experiments participants' STM was filled with material; e.g. a list of numbers -Participants had then to perform resource-demanding tasks; e.g. reasoning task =>If the STM store holds and processes information, then maximising load should impair the other task Performance was not affected as much as expected -The two processes don't depend on same memory component
Control Processes
Control the input of the short-term store -Filter the amount and quality of the information coming from the sensory registers =>attention Control how much information and which information is retrieved from the long-term store Select the type of cognitive operations that are applied to the information held in the short-term store
The Visuospatial Sketchpad (VSSP)
Deals with visual information Plays an important role in helping us keep track of where we are in relation to other objects as we move through our environment Stores and manipulates visual and spatial information held in LTM Capacity limited to around 4 items (Miyake et al., 2001) -Items are subject to decay
Forgetting in STM
Decay theory of forgetting =>Forgetting is due to decay - the fading over time of memory traces e.g. Peterson and Peterson's results Interference theory of forgetting =>Forgetting is due to interferences caused by neural events Two types of interferences =>Proactive interference = The influence of previously learnt items on the learning of new material =>Retroactive interference = The opposite influence from new items to already learnt items McGeoch and McDonald (1931) showed that interferences are likely to occur when new items are similar to items in memory, in particular semantically
The Phonological Loop (PL)
Dedicated to speech-based information Entry is automatic -Not under conscious control -Rehearsal under active control Input can be auditory or visual -Visual input is recoded into its phonological equivalent by orthographic to phonological translator; phoneme translation -Auditory input stored for a few seconds in phonological code -'inner voice' rehearsal processes (sub vocal) -Visual input is recoded into its phonological equivalent by orthographic to phonological translator Output is the spoken word
Word length effect
Ellis & Hennelly (1980) Study of English-Welsh bilinguals Digit names in Welsh have the same number of syllables as in English -But they are made of longer vowels Welsh speakers perform worse in the digit span task =>Welsh articulation rate 385 ms/digit =>English articulation rate 321 ms/digit Evidence suggests word-length effect depends on the duration of time taken to articulate the word -Not the number of syllables (Baddeley et al., 1975; 2000).
Stages involved in learning and memory
Encoding meaningful information; All three need to work in order for a memory
Murdock (1962) Free recall paradigm
Free recall paradigm -Participants had to remember lists of words -Immediate free recall Murdock found a serial recall curve, with primacy and recency effects Two different memory systems -Short-term and long-term memory Asymptote = a straight line that continually approaches a given curve but does not meet it at any finite distance.
Further reading
Gobet, Fernand, Chassy, Philippe, Bilalic, Merim (2011) Foundations of cognitive psychology, chapter 6
Irrelevant speech/Articulatory suppression
Hanley & Bakopoulou (2003) Examined the combined effects of irrelevant speech and articulatory suppression when target items were presented auditorily The effects of articulatory suppression and irrelevant speech were additive Suggesting that -Both affect the same component -Both affect rehearsal
Waugh & Norman's (1965) explanation
Information is put into short-term memory, where it is constantly rehearsed If information is rehearsed long enough, it is put in long-term memory =>Else, it is forgotten Displacement theory
Stroop task
Inhibit the desire to verbalise the colour
Shallice & Warrington, (1970)
KF, suffered a motorcycle accident causing damage to his left parietal occipital region -very poor digit span (<2 items) -still able to store new information STM -still retained 7/10 items some months later. STM -> LTM The multi-store model predicts that this should not be possible -It assumes that an intact STM is required to transfer information to LTM. Learnt a 10 word sequence in fewer trials than controls.
Sensory Registers
Keep a trace of sensations until the information is forwarded to the short-term store Hold modality-specific information for a few hundred milliseconds -Not accessible to consciousness Large memory capacity for sensory registers Large memory capacity for sensory registers - e.g. iconic = 12 items rather than 7+2 Sperling
CE not a unitary mechanism
Lack of correlation between CE tasks, neurological evidence and large scale experimental data suggests there are at least three separate functions (Miyake et al., 2000)
Free recall paradigm
One of eth oldest methods of studying memory but still one of eth most popular. Typical experiment has 3 phases; Talk about rehearsal during retention p93.
Criticisms of WM
Over-simplified, fails to reflect the complexity of "real" cognition (Andrade, 2001). Little is known about the central executive in terms of its capacity and functioning. =>how does the central executive integrate information from different modalities? Only involves STM so not a comprehensive model of memory Infers all spatial information is visual, but blind people have excellent spatial awareness (Lieberman, 1980) =>Does WM reflect general intelligence (Kyllonen & Christal (1990) or speed of processing (Salthouse, 1996) =>How does the Episodic buffer function?
Waugh and Norman (1965)
Participants are shown a list of digits -They are then given a probe; one of the digits in the list -Their task is to retrieve the digit following the probe =>E.g. List (9 0 8 5 3), Probe = 8, Answer = 5 Participants have to focus on one digit rather than the entire list -This technique essentially suppresses rehearsal Results -Weak memory for the last items -Suppressing rehearsal eliminates the recency effect
Episodic buffer summary
Passive integration system Combine codes from different modalities - e.g. Visual, phonological, LTM, perception - Semantic illusions a function of EB, rather than PL Still does not explain CE/rehearsal disruption equality Capacity of around 4 chunks Although initial scepticism raised, clearly there is an interactive link between LTM and WM - Episodic buffer explains this link - Evidence from patients with Alzheimer's disease
Baddeley & Hitch's Working Memory Model- Diagram (1974, 2003)
Questioned the concept of short-term store as proposed by Atkinson and Shiffrin's theory Phonological loop - holds speech based information. Visuo-spatial sketchpad spatial and visual coding
Serial Position Curve
Recency effect; good free recall of the last few items in a list based on information from the short-term store Primacy effect; a high level of recall of the first items in a list, mainly dependent upon extra rehearsal possible transfer to LTM - Retention interval = didn't affect primacy effect - LTM only recency effect - STM
Episodic buffer
Recent addition to the model (Baddeley, 2000, 2003) Essentially replaced the central executive for coordinating between the two slave systems Aims to explain that we perceive things as wholes and not as discrete parts -This integration of information in single percepts is done in the episodic buffer Temporary interface between slave systems, CE and LTM Is the place where the integrated information is held and is available to consciousness Multi modal Measured using the constrained sentence span
Evidence for STM & LTM
Scoville & Milner, (1957) -HM, a patient suffering from epilepsy, underwent surgery Ablation of parts of his right and left medial temporal lobe -Including about two thirds of the hippocampus and amygdala After surgery, HM was unable to store new memories into long-term memory But could perform tasks requiring short-term memory -e.g. holding a conversation HM had an intact STM but a damaged LTM Double dissociations -case studies and experimental
Evaluation of WM
Several advantages of WM over Atkinson & Shiffrin (1968) Short-term store WM is concerned with active processing AND transient storage of info; =>relevant to mental arithmetic, verbal reasoning, and comprehension. WM model overcomes many of the problems of the multi-store model, =>Ability to perform tasks together without disruption. =>Rehearsal an optional process The theory accounts well for the word-length effect, articulatory suppression, and the performance of various brain-damaged patients.
(Incorrect) assumptions of the Multi-store model
Short-term memory is a more complex system than proposed by Atkinson and Shiffrin; Oversimplified; STM found to be modality dependent; Focuses on structural rather than on processes.
Postman and Phillips (1965)
Tested the hypothesis of two memory systems Used a free recall paradigm with lists of 10, 20 and 30 words Two main conditions -No retention interval -Recall after a delay of 15 or 30 seconds, filled in with a counting task The presence of a retention interval led to the elimination of the recency effect
The Chunking Hypothesis
The cognitive system fuses together functionally-related bits of information; chunking (Miller, 1956) A chunk is -"a collection of elements having strong associations with one another, but weak associations with elements within other chunks" (Gobet et al., 2001, p. 236) -Chunking enables the 'fusion' of simple units into groups, can be used as units themselves Chunks are long-term memory structures that are learnt -Learning one chunk takes between 5 and 10 s (Simon, 1974) Learning to read - example of chunking letters of the alphabet together to form meaningful information. Capacity of STM should be measured in terms of chunks not individual items
Multi-store model of memory
The idea that memory is organised around sub-components - each fulfilling a specific function. Cognitive architecture. STM differs from LTM - capacity and interference.
Evaluation of Multi-store model
The model is influential; generated further research into memory. A cognitive architecture is proposed, with three major stages Input -> Cognitive processing -> Output Evidence to support distinction between STM and LTM: -Encoding, duration and capacity. -The model is supported by studies of amnesiacs: HM case study.
Central executive (CE)
The most important yet least well understood component of WM model. Baddeley equates CE to the supervisory attentional system (SAS, Norman and Shallice, 1980). -SAS is a limited capacity system and is used for a variety of purposes, including: -tasks involving planning or decision making -trouble shooting in situations in which the automatic processes appear to be running into difficulty -novel situations -dangerous or technically difficult situations -situations where strong habitual responses or temptations are involved Deals with cognitively demanding tasks -Dividing attention, switching attention, and focusing attention on task-relevant items -Functions of inhibition, updating and interference -Controls the information flow within and between slave systems =>Controls the information arriving from the senses and supervises retrieval from memory Domain general (Kane et al., 2007) FLEXIBLE, BUT FINITE RESOURCE
Subcomponents
The phonological store (inner ear): -Linked to speech perception. Holds information in speech-based form (i.e. spoken words) for 1-2 seconds. =>Information lost if not refreshed -Holds the items online in a passive way, in either an acoustic or a phonological format The rehearsal process (inner voice): -Linked to speech production. Used to rehearse and store verbal information from the phonological store. =>Articulatory rehearsal process -Limited capacity of 3-4 items -Sets items' activity back to its maximum, item by item Subcomponents cooperate to keep information for further processing Used heavily in tasks such as counting, reading
Levels of processing
The strength of memory traces depends on the amount of processing to which stimuli are subjected Deep processing of stimuli - stronger memory trace; easier to recall =>Elaborative rehearsal; stimuli are subjected to deeper, typically semantic processing that recode them more efficiently Shallow processing of stimuli - soon forgotten; difficult to recall =>Maintenance rehearsal; stimuli are simply rehearsed so that they can be held in short-term memory Only elaborative rehearsal leads to long-term learning
VSSP divided into two subcomponents (Logie, 1995)
The visual cache -Stores information about colour and shape The inner scribe -Deals with movement and spatial information -It rehearses information in the visual cache, transfers it to the central executive and is involved in the planning and execution of body movements. -Carries out cognitive operations on this information (e.g. rotating an object)
Aging and WM
Verhaeghen & Cerella, 2002 -Evidence of WM from age 3 years -WM matures after adolescence -In old age memory maintenance and the ability to switch between tasks declines -Sustained attention does not decline Schneider-Garces et al, 2010 More neurons activated to complete executive tasks in older populations.
Empirical Evidence for WM
What evidence is there that working memory exists, that it is made up of a number of parts, that it performs a number of different tasks? The working memory model makes the following two predictions: 1. If two tasks make use of the same component (of working memory), they cannot be performed successfully together. 2. If two tasks make use of different components, it should be possible to perform them as well as together as separately.
Outline and aims for today's lecture
What is memory? -Memory stores Short-term memory -Multi store model -Working memory model -Levels of Processing model By the end of the lecture you should; -Understand the complexity of memory =>and recognise the distinction between LTM & STM. -Be able to discuss classic theories of memory and evaluate methods used.