COGS 101B - Lecture 16 - Short-term and Working Memory

Sensory memory (fed by perception) seems to precede short-term memory, with duration of {0.5-3} seconds. {Unattended} information is lost.

Working memory has typical duration of {5-15} seconds. {Unrehearsed} information is lost.

Q. What’s the difference in conception between short-term memory and working memory?
A. Short-term memory was thought of as being for storage, whereas working memory is also used for manipulating information during complex tasks.

Short-term memory decay is driven by {interference} processes, not time alone. (Keppel and Underwood, 1962)

Q. How was the duration of short-term memory first characterized? (Peterson and Peterson, 1958)
A. Asking people to remember three-letter sequences while counting backwards from a high number (i.e. to prevent rehearsal)

Q. How did Keppel and Underwood (1962) revise the interpretation of Peterson and Peterson’s pioneering 1958 short-term memory study?
A. They observe that the first trial had very little forgetting, suggesting proactive interference: memories of previous trials cause the letters in subsequent trials to decay more rapidly.

The capacity of short-term memory depends on the {complexity} of the items being remembered, or in other words the amount of {information}. (Luck and Vogel, 1997; Alvarez and Cavanagh, 2004)

Elements of working memory short-term storage: {visuospatial sketchpad}, {episodic buffer}, {phonological loop}

{Phonological loop}: stores {verbal/auditory} information over periods of {a few seconds} by {rehearsing in a loop}

{Episodic buffer}: {holds and integrates} information from {phonological loop}, {visuospatial sketchpad}, and {long-term memory}. Creates a {coherent representation} of what’s in your conscious awareness.

Evidence for the phonological loop

{Phonological similarity effect}: in the context of {recalling rapidly flashed letters}, when {people made errors}, they tended to {report letters which sound similar}, suggesting involvement of {the phonological loop}. (Conrad and Hull, 1964)

{Word length effect}: in the context of {recalling rapidly flashed words}, people are {more likely to remember short words than long words}, perhaps because {they consume less of the phonological loop}. (Baddeley et al, 1984)

{Digit span} may be explained by capacity of {the phonological loop}: the number of words or letters you can remember is about as many as you could {say aloud} in {1.5-2 seconds}.

{Articulatory suppression}: repeating “the the the…” aloud causes {reduced memory of a word list}. (Baddeley et al, 1984)

The articulatory suppression effect eliminates {the word length effect}, which supports the idea that the latter relies on {the phonological loop}.

The prefrontal cortex is important to working memory

Monkeys who have had their {prefrontal cortex} removed can’t remember where a treat was after a short delay. (Goldman-Rakic, 1992)