Sensory and WM Flashcards
3 common operations of memory
- Encoding
- Transform sensory data into mental representation, can be called learning but doesn’t have to be that intentional, we can also encode things we are at least paying a little bit of attention to
- Storage
- Keep encoded information in memory
- Retrieval
Pulling info from memory to be used unconsciously or consciously
Understand basic terminology of memory research
explicit memory
recall vs recognition testing
sematic vs episodic (robot gradient)
pro vs retro active interference
explicit memory is the conscious and intentional encoding and storage of information it its measured by: recall, recognition tests also called direct tests
vebefits from deep and meaningful encoding of (levels of processing effects)
Recall tasks
- Producing an item from memory
Three main types:
- Serial recall, very strict e.g. phone number. Must be all in the right order
- Free recall, any order e.g. a shopping list. No order needed as long as all items are there
- Cued recall, information in environment used to trigger retrieval. e.g. paired-associates recall where cue could be anything.
Recognition
- Select an item which you have seen previously
- Can be true or false (old/new) tests or multiple choice
- recognition is usually better than recall because retrieval is cue dependant, the best cue to remember the item is the item itself
Semantic memory vs episodic memory
- smeantic knowledge is usually about other people, places or things, tends to be more crytalissed knowledge fact based
- episode is more personal, explicit to youand how you learnt the thing
- They are different memory systems yet the boundaries are unclean as Sematic memories used to be episodic but they have been decontextualised or (semantified)
- The evidence of them being separate systems is clear from the ribot gradient
- Ribot gradient: dementia patients lose recent episodic memories first
- Reversed Ribot gradient: in semantic dementia (rare) semantic knowledge and old episodic memories are forgotten first
Forgetting (from explicit memory)
Proactive interference (PI)
- Previous learning interferes with later learning
- Rest periods prevent interference
Retroactive interference (RI)
- Later learning interferes with previous learning
Rest periods prevent interference
Understand basic terminology of memory research
implicit memory
priming (semantic vs perceptual)
Implicit memory (non declarative)
- When we produce information from memory but were not aware were doing it
- measured using priming
priming testing
- Even thought you did not intentionally retrieve or consciously retrieve, seeing things more recently makes them more likely they will come to mind
Semantic
- E.g. stem completion
- More likely to choose a certain word because of recent exposure or semantic networks
Perceptual/repetition priming
- E.g. fragment completion, broken lined drawing > what animal is this?
- Repeated exposure speeds up processing
- Can be very long lasting
Other forms of implicit memory
- Procedural memory
- Conditioning
Testing implicit memory
- Often called indirect memory tests
illusion of truth effects: Familiar statements/products are more likely to be believed or trusted. Eg propaganda or which pasta sauce to use
Describe the Atkinson & Shiffrin model of memory
and its criticisms
- Three stores of memory all mainly one way
Sensory memory:
- capable of storying relatively limited amounts of information for limited periods of time (milliseconds)
- it either moves to STM or is forgotten
Short term memory:
- very limited capacity for somewhat long period (seconds), STM rehearsal leads to LTM storage
- is either forgotten or put int a response output or moves into LTM
- rehearsal makes it more likely to move into LTM
Long term memory:
- very large capacity for very long times (infinite)
- only via rehearsal
criticism
- Rehearsal is ineffective for LTM retention
- Patients with severe short-term memory deficits but intact long-term memory
- cannot explain chunking and the use of semantic knowledge to integrate info from STM to LTM
Cite evidence of criticisms against the Atkinson and shiffron model
STM isnt only gateway to LTM
- Patients with severe short-term memory
deficits but intact long-term memory (Baddeley &
Hitch, 1974)
LTM and STM arent uniform processes
- Non-unitary short-term memory deficits (e.g.,
Patient KF, Warrington & Shallice, 1972)
* Forgetting of auditory stimuli greater than visual
stimuli
* Forgetting of verbal materials such as letters,
words and digits, not other meaningful sounds
(e.g., animal sounds, ringtones)
Describe the main characteristics of sensory memory (iconic and echoic memory)
Sensory memory
- Simultaneous entry of sensory information into consciousness
- Relatively limited capacity, very brief duration
- Two main types: iconic and echoic
- Both spontaneously decay and are erased by masks or inference
- function of it it to let us interact w stimulus even after its gone eg for reading or listening comprehension
Iconic sensory memory
- Visual
- Discrete visual sensory register that holds information for a short time like a snapshot
- Info is stored as visual images
- Some argue its purely visual/perceptual and therefore no semantic processing
- hold or 9 items for 500 m/s
Echoic sensory memory
- Auditory equivalent of iconic memory
- Brief, unanalysed memory of an auditory stimulus (an ‘echo’)
- hold 3 items for 250 m/s
these have been said to be an underestimation
Explain how studies have investigated the capacity and duration of sensory memory
Echoic
to investigate this:
Massaro (1970)
- presented people ith a tone and they just have to say whether it high or low
- then they are presented with masking tone, this is either delayed or immediate
- when 0ms delay 60% got it correct
- when 250ms delay 90% correct as its not interfering w original tone
Explain how studies have investigated the capacity and duration of sensory memory
Iconic
to investigate the capacity of iconic memory:
Sperling (1960) – Full Report Procedure
- presented 4-12 items all at once for 50ms total eg Y H 7 J
- then asked to report as man as they saw
- demonstrated its capacity to recall is 4-5 items
- can tell you there was 12 in total but only recall 4
- underestimates capacity
Sperling (1960) - partial report procedure
- presented 4-12 items all at once for 50ms
- a tone is presented
- if its high, medium, low tone report top, middle and bottom row respectively
- participants typically report 3 items but the whole array must be encoded to do this
- therefore the total capacity is around 9 items
- rapid decay if the tone is delayed
Describe the Baddeley multi-component working memory model
Baddely and hitsch model
- is a central pool of attention model
- first only the central executive, phonological loop and visuospatial sketchpad were in the model
- then crystallised systems were added in 2000
- then episodic buffer
Central Executive
- the central master
○ modality-free attentional control system
○ resolves conflicts over what cognitive processes should occur next
○ coordinates information from various sources
○ no storage capacity of its own
- Single pool of attention in charge of two stores
visuospatial sketchpad
- responsible for the storage of visual and spatial information
- mode specific
phonological loop
- responsible for the storage of verbal and acoustic information
- If you’re not rehearsing something phonological loop it will decay
mode specific
crystalised systems
- visual semantics
- language
- episodic LTM
- ‘Visual semantics’ and ‘language’ is semantic long-term memory
modules that interact with WM
- explains why it is easier to hold/chunk items in WM that have semantic meaning
episodic buffer
○ Controlled by central executive
○ Limited-capacity storage system capable of integrating information from a variety of sources
○ Serves as an interface between systems
- Holds “episodes” of integrated, multimodal information
- explains visual effects on verbal recall
- explain ineractions between LTM and WM eg:
retrieve unsuccessful attempts to solve a problem, in order to modify your strategy or serves long-term episodic ‘learning’, binding the contents of WM into coherent ‘episodes’, which are then stored in episodic LTM.
how to identify which component of the baddely and hitch model is being taxed (robins et all)
used these secondary tasks during a chessgame
the central executive
- random number generation task
- performance decrements
the phonological loop
- articulatory impression
- same as controls
visuospatial sketchpad
- spatial tapping
- performance decrements
Baddely and Hitch WM
Cite empirical evidence that has supported its development
evidence that the loop and sketchpad are components which work separately but the central executive is in charge
quinn and mconnell (1996)
- Learn list of words using 2 conditions: visual imagery (visual) or rote rehearsal (verbal)
- 2 interference conditions: dynamic visual noise (visual) or irrelevant speech (verbal)
- the interferences was during the learning task
- fount a cross over interaction
- if the conditions were from the same component oeg both visual then there were performance decrements whereas if they were from separate components we saw better accuracy comparatively
- verbal interferes w verbal and visual interferes w visual
evidence for decay in the absence of rehearsal
Phonological similarity effect: P G T C < R G F S
- Speech-based rehearsal: similar items interfere with each
other
- Articulation rate effects (e.g., word length): Better memory for short vs. long words (e.g., ball, plum >
tricycle, pineapple, or even tick, goal > grass, moon). More rehearsal = less effect of decay
- Articulatory suppression effects (“the, the, the”): Memory performance declines, Prevents rehearsal = more effect of decay
ADDED CRYTSALISED LTM
evidence that LTM and WM interact
- evidence from patients with impaired
phonological STM, but intact phonological LTM
(for arbitrary sequences of words, cf. Shallice &
Warrington, 1970). these patients do have difficulty with long-term
phonological learning with unfamiliar material such as a new language
- Phonological WM performance predicts
vocabulary learning in children (Gathercole et al.,
1992)
ADDED EPISODIC BUFFER
evidence that visual and phonological info are combined
- Effects of visual similarity on STM for verbal materials so words that sounds the same AND look the same are easier to remember than words which sound same look diff e.g., fry, cry, dry < guy, sigh, lie (Logie et al., 2000)
evidence that long term crystalised and WM interact
- Effects of meaning on WM performance: Series of pictures or words easier to remember when they make a story or sentence (e.g., Baddeley et al., 1987). also explains chunking
