Metacognition

Metacognition

knowing that you know, tip of the tongue phenomena

Monitoring Metacognition

making judgments about one’s memory/cognition

Control Metacognition

using those judgments to guide behavior

Kornell, Son, and Terrace - Experiment 1 (monkeys)

method: DMS task - monkeys sees series of six slides, each with one picture in a different position - then sees picture with 9 potential pictures - must choose which of the 9 stimuli it saw in the sample of 6 - then monkey indicates how confident it is about the choice - high confidence, correct (+3)/incorrect(-3) - low confidence (guaranteed 1)

results: when monkeys were performing well, they were pressing the high confidence icon and opposite for bad, they knew how confident they were so proof of metacgonition

Kornell, Son, and Terrace - Experiment 2 (monkeys)

method: standard 4 item serial order task - training on a 50% hint condition

results: initially, animals used the hint icon all the time, over time the animal became less reliant on the hint icon as they learned the order

Hampton (2001) Flag Study (monkeys)

method: visual DMS - monkey sees stimulus - short delay - 33% it is a forced trial/67% it is a free trial - in forced, animal must take test - in free, animal can choose to take test or bail

hypothesis: monkey should do better in the test on the free condition than the forced condition

results - overall accuracy: both monkeys performed significantly better in the free trial

results - no sample condition: if you don’t get the sample stimulus you should bail, only one monkey consistently declines the test, other one still does decline far more than normal

results - increasing the delay: increasing the delay should result in poorer performance because memory gets worse, thus animal should bail more - monkeys performance aligns with predictions, always decline the test at the longest delay period

Inman & Shettleworth - Experiment 1 (pigeons)

method: pigeon sees sample stimuli - retention interval of 1,2,4,8 seconds - pigeon given one of three conditions (safe key, memory only, combined) - if correct they get 6 pellets or 0 if wrong, safe key is 3

results: for most birds, combined was better than memory only condition, the selection of the safe key doesn’t really increase across the delay periods like it does for monkeys

Inman & Shettleworth - Experiment 2 (pigeons)

method: similar to Hampton’s flag study where they must first choose what they want to do before they are actually shown the condition

results: no difference between memory only and combined, safe key goes up a little bit but not significantly

Inman concludes that there is no evidence for metacognition in either experiment

She argues that the pigeons do better in experiment 1 because of signal detection theory because the reward structure is different than in experiment 2