PSY370H1 Lecture : LEC4 – Insight and the Problem of Transfer Oct 1 2009
PSY370: Thinking and Reasoning
LEC4 – Insight & the problem of Transfer
Oct, 1st, 2009
Metcalfe 1986 (a)
! meta-cognition
! abilities to monitor & intervene in one's cognition
! one's ability to predict future ability to remember
! research shows that this is a reliable faculty
! ie. when ppl predict their aptitude at this ability they are pretty accurate
! people have feelings of knowing (FOK)
! if insight problem solving involve standard memory search as per W & A, and if memory
searches are positively and reliably correlated with meta-cognition (FOK), then insight problem
solving should be positively and reliability correlated with meta-cognition (FOK)
! study 1
! compared relationships b/w FOK and how they actually remember or whether or not they
are going to solve an insight problem
! FOK correlates high to remembering but low to success at insight task
! thus this finding is hard to reconcile this finding with W & A
! but the experiment did NOT show there was no memory search in insight problem solving,
it does NOT prove that a gradual search is not in process
! it does indicate that the relationship b/w memory and meta-cognition is somehow altered
when doing insight problem solving
! but ppl were more confident with regards to insight task than it was for the memory search
task even though the success rate was reversed
thus confidence does not reliably track accuracy
! Study 2
! wanted to refine this probe
! looked at GPS & Simon
! Simon: Means-End analysis is dependent on noting salient differences heuristic
examination of the past leads to feelings of whether you are feeling warmer/colder wrt
goal
Simon found evidence of ppl using this warmth heuristic & it was a good predictor of
success
this is a way to avoid combinatorial explosion
! FOW: Feeling of Warmth
fine grained
! had subjects try to solve insight problems but every 10 seconds subjects have to provide
warmth rating
! Result:
1. abrupt feeling of warmth rating right before solutions are found
2. strong feelings of warmth ratings actually predicted failure to solve the problem
subjects satisficing rather than solving the problem: they try to turn it into simpler more
familiar problem
when high warmth is correlated with failure then the graph is incremental in nature, but
when it is correlated with success then the graph is abrupt
! Study 3 Metcalfe & Wiebe 1987
! 2 sets of problems: insight + non-insight
! 2 specific predictions
www.notesolution.com
Document Summary
Lec4 insight & the problem of transfer. abilities to monitor & intervene in one"s cognition. one"s ability to predict future ability to remember. research shows that this is a reliable faculty ie. when ppl predict their aptitude at this ability they are pretty accurate. compared relationships b/w fok and how they actually remember or whether or not they are going to solve an insight problem. fok correlates high to remembering but low to success at insight task. but ppl were more confident with regards to insight task than it was for the memory search task even though the success rate was reversed thus confidence does not reliably track accuracy. simon: means-end analysis is dependent on noting salient differences heuristic looked at gps & simon examination of the past leads to feelings of whether you are feeling warmer/colder wrt goal.