Quantifying Memory Lecture Notes

Vocabulary flashcards defining direct and indirect memory test types, accuracy measures, response bias metrics, and qualitative performance measures.

Direct memory test

A memory test in which the subject is asked to consult his/her memory of the studied material.

Free recall

A direct memory test where the subject is asked to say or write all the items they remember from the list they just studied.

Cued-recall

A direct memory test where the subject is asked to say or write items from the studied list using a specific prompt or cue (such as category names or paired words).

Forced-choice recognition

A recognition test format where a subject is presented with options and asked to pick the item they studied.

Old/New recognition

A direct memory test (also called Yes/No recognition) where the subject is presented with an item and asked whether it was on the study list ("old") or not ("new").

Chance performance

The score that a person with no memory would get on a test just by guessing, calculated numerically as 1 / number of response alternatives.

Response bias

The tendency for a participant to favor one particular response option when guessing on a recognition test.

Liberal bias

A response bias in recognition tests where a participant tends to say "yes, old" when guessing, resulting in a response bias score greater than $0.50$.

Conservative bias

A response bias in recognition tests where a participant tends to say "no, new" when guessing, resulting in a response bias score less than $0.50$.

Hit

An outcome in an Old/New recognition test where an actual Old item is correctly identified as "old".

Correct rejection

An outcome in an Old/New recognition test where an actual New item is correctly identified as "new".

Miss

An outcome in an Old/New recognition test where an actual Old item is incorrectly identified as "new".

False alarm

An outcome in an Old/New recognition test where an actual New item is incorrectly identified as "old".

Hit rate

The proportion of Old items correctly identified, calculated as the number of hits / total number of Old items.

False alarm rate

The proportion of New items incorrectly called old, calculated as the number of false alarms / total number of New items.

Accuracy score

A measure of accuracy in Old/New recognition tests computed by (False alarm rate - Hit rate) to exclude response bias.

Response bias score

A metric calculated as False alarm rate / $(100 - \text{accuracy})$, representing the percentage of errors that were of the false alarm type (ranging from $0.00$ to $1.00$).

Response confidence

A quantitative memory measure that is normally lower for errors than correct responses; high in errors indicates an anomaly requiring explanation.

Semantic clustering

A phenomenon during free recall where a participant recalls items grouped in order according to their semantic categories.

Intrusion

An error in a free recall test where a subject recalls an item that was not present on the original study list.

Indirect memory test

A memory test where the subject is not asked to consult his/her memory, but performance is nonetheless influenced by prior experience.

Word-stem completion

An indirect memory test where participants are asked to complete word stems (e.g., CHA _ _ _) with the first word that comes to mind.

Baseline performance

The measured rate at which subjects produce target words on an indirect memory test without having seen them in a study phase.

Priming score

The memory score on an indirect memory test, calculated as hit rate after study - baseline rate.