Theme 1: Short-term memory Flashcards
phonological similarity effect
similar sounding letters/words are remembered worse than differently sounding ones
Baddeley & Hitch: Three-component model of WM
Central executive, visuospatial sketchpad (visual cache, inner scribe) and phonological loop (phonological store, articulatory rehearsal process)
How is serial order maintained?
chaining interpretations: each item forms a cue/stimulus for the following item
problems with serial order: what is the same item reoccurs? what if items are phonologically similar?
- Burgess & Hitch: contextual models: order is maintained through contextual cues
- page & norris: its based on positional associations between the first and sebsequent items
proposed function of the phonological loop
facilitate acquisition of language (phonological loop capacity is good predictor of acquisition of new language)
episodic buffer
- limited capacity store binding together episodic information
- connected to long term episodic memory
- attentionally controlled by the central executive
N-back test
subjects are presented with a continuous stream of items and are instructed to press a key when they detect a repetition at a specified delay (for example: n = 3, press key when repetition with three items ago)
Norman & Shallice’s model of attentional control
attentional control is divided into 2 processes:
- control of behaviour by habit patterns and schemas
- attentionally limited controller called the supervisory activating system (SAS) = intervene when routine control is insufficient
fluid intelligence
ability to reason and solve problems independent of previously acquired knowledge
Jaeggi article: improving fluid intelligence by WM training+ limitations
Goal: does WM training lead to transfer to fluid intelligence?
Method: N-back test (auditory string of letters or spatial locations on a screen), as performance improved –> n also improved
Results: exp & control group both improved, but exp group improved more
limitations:
- no threshold phenomenon: how long could continue to improve?
- no longitudinal effects: long term effect?
Harrisson et al: WM training may increase WM capacity but not fluid intelligence
-Criticism on Jaeggi: complex span task & n-back does not measure same construct
-goal: does WM training improve WM capacity? does that transfer to fluid intelligence?
-hypotheses:
–Near transfer: WMC improved similar task/stimuli
–moderate transfer: different tasks, both measuring WMC
–far transfer: complex span task –> improved in fluid intelligence
-results: moderate transfer for complex span and simple span training groups
how do we use STM when we need to retrieve information? Serial vs parallel vs cascading
-Serial = search through items one by one until we find our answer
-parallel = multiple items are held own STM at the same time & divided among cognitive resources
-cascading = mixed both
Sternberg investigated r theories of STM search:
- Parallel search: all items are addressed at the same time –> RT does NOT vary with set size & no difference between yes&no responses
- Serial self terminating search: serial then stop when you find the item –> increase in RT with set size & steep slope for no & yes, but yes is half the size of no
- serial exhaustive search: always searching the full set serially –> increasing RTs &no difference for yes/no answers
RESULTS: support for serial exhaustive search