Transcript Brown-Peterson task

Short Term Memory
•
•
•
William James (1900): Made Primary-Secondary memory
distinction. Important distinction: Primary was limited
capacity and duration. Secondary more permanent.
George Miller: Magical #7 (1956). Emphasized the
importance of recoding or chunking of information.
Brown-Peterson task: test of duration of STM
Trigram: KNP; 517; backwards by 3 from number for variable
amount of time, by 15-18 seconds trigram gone
Decay or Interference?
•
•
Brown-Peterson studies suggest that it might be delay that leads to loss in STM
Waugh and Norman disagreed, argued that verbal repetition of numbers may have
constituted interference effects.
• Waugh/Norman (1965) study: decay vs. interference.
List of numbers presented: 7 9 5 1 2 9 3 8 6 4 3 7 2 (tone)
Tone marks repeated number; must recall number coming after repeated number first time
(answer: 9)
Two variables: rate of presentation: (1/sec; 4/sec)
Number of intervening items (1-13)
If decay then rate should be critical
If interference then number of items
No effect of rate; significant effect of items.
Conclusion: Far stronger effect of interference
than decay
Two forms of interference:
Retroactive: newer blocks older (present in WN study)
Proactive: older blocks newer
Wickens Release from PI:
forgetting due to build up of semantic proactive interference
items– distracter-recall; same category items -distracter-recall; same category
items distracter-recall; then switch (for experimental group).
STM/LTM distinction: Serial Order Effect
Primacy: rehearsal allows for
encoding into LTM
Recency: still present in STM
Delay before recall affects
recency but not primacy
part of the effect
Searching STM: Sternberg Paradigm
Sternberg Task
Search set (1-6 items); Comparison item; yes/no part of set?
Ex: K X G D (G=yes) (F=no)
RT’s increase linearly with set size; RT for yes and no equal. Serial/Exhaustive search.
Baddeley’s Model of Working memory
PL: phonological store + articulatory loop
Articulatory suppression effect: talking
suppresses memory
Phono similarity effect: similar sounding
words harder to remember than dissimilar
Brain anatomy of WM:
Dorsolateral Pre-frontal Cortex: CE
Broca’s area and left parietal: PL
Occipital and posterior parietal: VSSP
VSSP: Mental rotation studies
Farther apart pairs are from similar orientation
the longer the reaction time. Longer rotations
take more time
Episodic buffer: putting it all together
•
•
•
•
•
•
EB: binds info from other systems together to form meaningful memories.
Dual task robs resources from non-meaningful processing not meaningful
Jefferies, Lambdon, Ralph and Baddeley (2004)
Subs learned unrelated words or sentences or meaningful sentences that formed a story
When secondary task applied (monitoring computer screen for form), performance on nonmeaningful conditions declined, but not on meaningful sentence condition
Ability of EB to use meaning for bind memories allowed for excess resources in meaningful
sentence condition, but resources in non-meaningful conditions all used up trying to recode
word/sentences.
Selective interference: Dual task studies in WM
•
•
•
Rationale: if PL and VSSP are truly separate systems then loading two tasks into
one should cause a greater decrement in performance than spreading two tasks
across two systems (one in each).
Task 1: either letter span (PL) or image span (VSSP)
Task 2: either adding (PL) or imaging (VSSP)
Measuring WM capacity: WM span
Requires both processing and storage
Ex:
6x2-2=10? SPOT
5x3-2=12? TRAIL
3x2-4=2? BRAND
Correct response: y,n,y SPOT, TRAIL, BRAND WMS=3
Increased WMS associated with:
Less distractibility: less likely to hear name in unattended channel in dichotic listening test
Better filtering of irrelevant details when processing text
More consistent responses on moral reasoning tests (consistent moral principle applied across different
dilemmas for high span subjects)
Word fluency: better able to retrieve category-relevant words from memory