structure of Memory Pt 2 Flashcards
Recall
Free recall: Tell me everything you can remember
-span task: experimenter reads out numbers + ppt have to recall
cued recall
-paired associates test: exp tells them pairs + ask to recall the word associated with presented word
Recognition
-exposure phase -> expose to stimuli
-Test phase- have old + new ones + ppts have to recognise which are which
-If memory of new items is weak + strong for old there is a distinction
-more likely for there to be an overlap - > some old remembered but some not, some new recognised but some not
-WM performance isn’t perfect
-D’ (d prime) = the distance between old + new distributions reflect: item familiarity, memory performaNE
Decision Making
-ppts have response bias
-ppts can guess
- c is bias
- Any items stronger than C are remembered
-Any items less strong than C aren’t remembered
- some ppts have different c
placements
- cautious = more likely to say its a new item
-Liberal = more likely to say its an old item
- can count the correctly recognised + false alarms ( believe seen before)
- closer Dis to 1 the better the WM
B + decision making
-reflects the placement of the criterion on the x axis
-reflect biases associated with the response
Cowans k value
- cowan 2011 et al
- change detection paradigm
- ppts have to remember orientation of stimuli
- get test stimuli + ask if lines are in correct orientation as before
capacity during development
- increase in verbal-storage, complex memory span + visuo-spatial memory as age increased (Gathercole et al 2004)
-At old age WM decreases (McNab
chunking + LTM
- Not number of items but chunks of info
- Miller- WM can hold 7 +/- 2 chunks of info
Time slot model
-Have a fixed amount of slots + put chunks in slots, once its full can’t fit anymore
- No evidence to support this
- precision not capacity
Mixture model
- Bays, catalao + Hussain 2009
- Target = red
- Non-target = blue
- Target distribution
- Mis-binding = non target distribution
- guessing = a constant -even distribution of responses across colours
-More precision = higher number - increased number of items to remember
- found smooth decline so cant be slot model must be resource model = amout of memory given to each item decreases but each item does have memory assigned
chunking + LTM
- Not number of items but chunks of info influence LTM
-Miller 7+/-2 chunks
Time slot model
-Have a fixed amount of slots once full can’t fit anymore chunks
- No evidence that supports this
- precision not capacity
Mixture model
- Bays, catalao + Husain 2009
- Target distribution
- Mis-binding = non target distribution
- Gussing = even distribution of responses across colours
- more precise = higher number
comparing resource vs slot model
- Bays + Husain 2008
- increased number of items to remember
- found a smooth decline so isn’t slot model
- resource model = amount of memory given to each item decreases but each item does have memory assigned to it
Attention + resource model
-ppts had to move eyes to item as snown on screen+ remember where it was
-measures precision
- Last item has much more memory as they could allocate the WM resource flexibly
- Allocation of WM resource can be bias by selective attention
- WM as a flexible resource
LTM - Retrieval cues
- Ability to retrieve memory of holiday from lots of retrieval cues because there is a network of associations between retrieval cues
-when 1 thing in the network is activated it spreads to associated things = retrieval
Paivo - Dual- coding Hypothesis
- for imageable words there are 2 routes to retrieval
- 2 concrete things can be remembered by imagining them as interacting
Encoding specificity/ transfer appropriate processing
-retrieval cues will be more effective the more similar they are to the conditioned present at encoding
-Tulving + Thomason ->
-presented with chair-glue stimulus
- one way to retrieve chair is to present glue
- Another way is to present table (word AsSoC)
-works better when present glue
- using table is dissimilar to the encoding of glue + chair
cortical reinstatement + encoding
-retrieval involves the same pattern of brain activity present when the memory was encoded
- Multivoxel pattern anAlysis technique - pattern of brain activity
- Wagner et al 2013
Declarative VS Nondeclarative
- declarative/explicity memory = memories consciously recalled consisting of info explicitly stored + retrieved
-Nondeclarative /implicity memory - memories not consciously recalled, implicitly stored info
-saw this in HM - believed never done task before but performance increased because he had implicit memory
Implicit memory
- conditioning - Pavlov
- dog salivating at bell is believed to be implicit memory
implicit Memory - Priming
Warrington + Weiskrantz
- shown word then showed them degraded/faded words
-If ppts had seen word before they are better at recognising
-Amnesic patients
- performed badly with recognition procedure (explicit)
-performed normally when shown visually degraded versions (implicit)
Implicit Memory- skills
-Masters 1992
-ppts learn how to play golf
- concurrent task given explicit
- Anything learned has implicit
- told them being eval = stress provoking
-explicit learning affected by stress -performed worse
- implicit learning more resistant + performance improved
Declarative memory - episodic vS semantic
-episodic = single episodes/events
- semantic = stores accumulative knowledge about world
-semantic better if there is episodic element to it
Long-Term Memory summary
- declarative or nondeclarative
- declarative can be semantic or episodic
