Metamemory Flashcards
Explain Nelson and Narens (1990) model for metacogntive processes in terms of memory
Object level processes, such as memory, are monitored by the meta level, with this monitoring and other top-down input like beliefs, the meta level controls the object level
What is a tip of the tongue state? Give a study as evidence
William James describes tip of the tongue states as an active gap that we can sense the shape of it - we feel close to remembering something, know what it isn’t, but can’t quite reach it. Yaniv and Meyer (1987) induced tip of the tongue states by asking for rare words from their definition, people sensed they knew the answer but were unable to recall it.
What is feeling of knowing? What is a FOK judgement rating?
Feeling of knowing is the sense that you know something, whether you can recall it or not. FOK judgements are formalised in research by participants rating how strong their feeling of knowing is - this can be a yes or no or on a numerical scale. FOK judgements are often prospective e.g. would you recognise the answer if you saw it?
What method did Hart 1965 introduce and what was found?
Recall, judge, recognise method (RJR). People asked general knowledge questions and attempt to recall, they rate their FOK judgement, then recognition of the answer is tested with whether they pick it from multiple choice options. Found that there was better recognition with a FOK (66%) than with no FOK (38%). FOK judgements weren’t perfect though as the strongest FOK judgement only had 75% recognition. Shows that we can make metamemory judgements better than chance but not perfect.
What are the two ways we can measure the accuracy of metacognitive judgements?
Calibration = the absolute accuracy, degree to which FOK judgement aligns with or deviates from real performance (recognition). Resolution = relative accuracy, whether the difference in accuracy across questions correlate with the difference in FOK.
What are the suggested ways we make FOK judgements? Which is better and why?
Direct access Hypothesis versus heuristic cues/inferential knowledge. Second better as FOKs are not perfectly reliable and can be very fast, and direct access doesn’t work conceptually if memory stored as synapse patterns.
What heuristic cues are used to make FOK?
Familiarity of question (e.g. whether feel have heard of the concepts/content before) and accesibility (partial retrieval cues e.g. starting letter in tip of tongue state).
What did Reder 1987 find showing we use familiarity to make FOK judgements?
When primed with words that appeared in game show style task questions, 7% increase in feeling of knowing (quick button press) also 4% drop in accuracy. Shows that FOK driven by familiarity regardless of whether know answer or not.
What did Koriat 1993 find showing we use accesibility to make FOK judgements?
Used RJR paradigm with consonant strings e.g., FDKR, then distractor task, asked to retrieve (but not guess) and make FOK judgement. Found that greater FOK with more correct letters remembered, but also greater with incorrect letters remembered. Shows that FOK judgement depends on whether information comes to mind regardless of whether correct or not.
What did Costermans et al. 1982 find showing we use inferential knowledge to make FOK judgements? What about Atir et al. 2015?
FOKs are a weak predictor of succesful recognition but correlate strongly with judgements of familairty and whether others would know. Atir et al. found experts overclaim impossible knowledge (90% say know something about made up term). Both show FOK comes from beliefs about what people generally know/should know.
What did Kolers and Panef 1976 find about using familarity in retrieval? What about Reder and Ritter 1992?
Fast knowing not decisions, quick to say don’t know something - fast judgement based on heuristic saves effort. Reder and Ritter found when given choice to retrieve or caclulate arithmetic question, people attempt to retrieve more if have seen the numbers in the question before, even if sign is different. Both show retrieval doesn’t start instantly but after a quick metacogntive judgement.
What did Costermans et al. 1992 find about using accesibility in retrieval?
Time trying to retrieve increases with greater FOK. Can use FOK judgements to control how long try to retrieve.
What did Koriat and Levy-Sadot 2001 find about using familarity and accesibility in retreival?
Asked general knowledge questions seperately manipulating familarity e.g., whether likely heard of more and accesibility e.g., whether likely to know. Found interaction - if unfamilliar then accesibility doesn’t matter as don’t try and retrieve, when familair try to retrieve so accesibility matters. Dynamically control retrieval process with familarity then accesibility.
What did Costermans et al. 1992 find about using confidence in retrieval? What about Koriat and Goldsmith 1996?
Time trying to retrieve decreases with greater confidence. Koriat and Goldsmith found when given choice to try and recall answer or pass, more likely to try and recall when confident in what recalled previously. More correct answers 82% when given choice to recall or pass compared to when no choice 47% - metacognitive judgement helps accuracy.
Where do accesibility, familarity, FOK, confidence, inferential knowledge and beliefs fit into Nelson and Narens 1990 model?
Inferential knowledge and beliefs feed top down into meta level. Feeling of knowing and confidence come from meta level, shaped by accesibility and familarity at object level.
