Cognition Chapter 5
The Route into Memory
- For many years, theorizing in cog-psych was guided by information processing – complex mental
events involve a large number of discrete steps, in which each step provides an output, used as the
input for the next step in the sequence.
Within this approach, a lot of the theorizing focused on acquisition of information (how info is
perceived & then moved into memory storage).
- First model by Waugh & Norman, refinements added by Atkinson & Shiffrin = the modal model.
Updating the Modal Model
- Modal model: information processing involves different kinds of memory. When information first
arrives, it is stored briefly in sensory memory, which holds onto the input in “raw” sensory form (an
iconic memory for visual inputs & an echoic memory for auditory). A process of selection &
interpretation then moves the information into short-term memory (the place you hold info while
you’re working on it). Some of the information is then transferred into long-term memory, a much
larger & more permanent storage place.
Modern theories: sensory memory (brief storage) plays a much smaller role
Also, working memory is used instead of short-term memory, to emphasize function.
Ideas/thoughts here are currently activated & being thought about (currently being worked on).
- Long-term memory (LTM) contains all of your knowledge & beliefs, mostly not being thought about
at the moment.
- The OG model describes working memory as a storage place/loading dock just outside of the LTM
“warehouse” – meaning it must pass through the working memory to get to LTM & transfer from
LTM back to working memory during retrieval.
Modern theories: working memory is a status, not a place – “in working memory” meanings that
ideas are currently activated & being worked on by a specific set of operations.
- Working memory is limited in size, long-term is vast.
- Getting info into working memory is easy (thinking about something = in your working
memory). Getting info into LTM involves some work.
- Getting info out of working memory is also easy (holding the ideas you are working on right
now, so the info is readily available to you. Finding info in the LTM can be effortful, slow &
may even fail. - The contents of working memory can be fragile, because they contain the ideas you are
currently thinking about, which could easily be pushed out. In LTM, info remains in storage
regardless (less fragile).
Working Memory & LTM: One Memory or Two?
- Free recall procedure: participants listen to a list of 30ish words each second & must repeat them
back in any order.
People usually remember 12-15 words: extremely likely to remember the first few words (primacy
effect), & the last few words (recency effect). Results in a U-shaped pattern describing the relaton
between position within the series (serial position) & likelihood of recall.
Explanation: working memory can only hold 5 or 6 words, so as the hear the list, earlier words will
continually be pushed out by later words. This explains recency effect. As participants read the list,
their goal is to memorize it, so they tend to repeat the first word over & over again to themselves –
which is called memory rehearsal. This occurs for the first few words. Note that first word has
advantage because it has 100% of your attention, when the second word arrives, it only gets 50%, the
3 must then split your attention with the first 2 words (33%). This explains the primacy effect; earlier
words receive a memory advantage (more time for rehearsal – greater chance of being transferred to
- These claims support the notion that recency is a result of working memory and primacy is a result of
- To test: After the list, participants must count backward from 201 by 3’s for 30 seconds. Counting
backward results in a loss of the recency effect because the numbers now replace the last few words in
the working memory, but has no influence on primacy effect because it has to do with the LTM. A
short delay with no interpolated activity does not interfere.
- Slowing the list presentation improves memory of pre-recency items, but does not improve words
initially effected by the recency effect.
Recency & pre-recency items on the list are impacted by different sets of influences, indicating that
these portions of the curve are products of different mechanisms. fMRI scans also show that memory
for early items on the list depend on brain areas associated with LTM.
A Closer Look at Working Memory
Function of Working Memory
- Sometimes we need to hold multiple ideas in our thoughts simultaneously.
- People differ in their “holding capacity”. The procedure used to measure this capacity has changed
over the years.
- For years, the holding capacity of working memory was measured with a digit-span task (read a
series of digits & immediately repeat them back – given longer lists until the person starts to make
errors). The number of digits a person could repeat without error is their digit span.
Using this test, working memory holds on average “7 plus-or-minus 2” items.
- Refinement required: what is an “item”?
- Miller: working memory holds 7 plus-or-minus 2 packages, & what those packages contain is largely
up to the person. If a person reads a series of numbers & thinks of them as individual letters, then they will
only remember around 7 of them. If the person “chunks” the letters into 3s/syllables, they will
rememer around 7 of these “chunks”, & therefore many more letters.
HOWEVER, the some effort is required to “repackage” the materials (assemble) & therefore
less attention is available for rehearsing.
- Regardless, chunking can create flexibility in what the working memory can hold.
- e.g. one individual who thought of number sequences as finishing times for races (3 min & 49.2
seconds). These can be grouped into larger chunks (times grouped into heats in a track meet). Using
this method, this individual increased his memory span from 7 digits to 79 digits! * However, this
person has only changed his chunking stategy, not the memory itself!
- Working memory is more of a status than a place – there is no specific place within the brain, instead
working memory is associated with a wide range of brain sites dependent upon what is being worked
- Memory’s capacity is therefore measured in terms of operation span – which is measured in several
- e.g. Reading span: Asked to read 2 sentences out loud & the repeat the last word of each sentence. If
they accomplish this, they must do the same with 3 sentences, & so forth until the performance limit is
located – known as the person’s working-memory capacity (WMC).
This task involves storing material for later use while simultaneously working with other
materials – exactly what working memory must do.
- The hypothesis then, is that people with a larger span (greater WMC) should do better in tasks
requiring the coordination of different pieces of information. This is found to be true: people with a
great WMC have an advantage in reasoning, reading comprehension, standardized academic tests, etc.
Also seem to keep their thoughts “on task” better.
- Therefore: helpful to have a larger WMC, link between WMC & intellectual performance give us an
idea of what we’re measuring with “intelligence” tests, & dynamic measures (operation span) should
be used over the span test (more static).
The Working-Memory System
- Previously described working memory as a system, at the center of which lie a set of executive
processes that govern the selection/sequence of our thoughts (named the central executive – but actuall
just a set of mechanisms & processes).
- The central executive is needed for the “work” in working memory, but there are settings in which
you need to keep ideas in mind – not to analyze – but because you’ll need them soon, in which case
the executive is not needed. In this case you only need the executive’s “helpers”, leaving the executive
free to work on more difficult matters.
- Helpers (compared to piece of scrap paper): visuo-spatial buffer (store visual materials), articulatory
rehearsal loop (store verbal material).
e.g. to multiply 12 by 23, you’ll likely multiple 12 by 20, then 12 by 3 & then add these products
together (3 steps). Between the first & second step, you were required to hold onto the first product
while you calculated the second. To do this, you shuffle the first product off to the helpers (rehearsal
loop), allowing the executive to do the new equation. Then for step 3, you can read the content of the
loop (look at your scrap paper) & finish the calculation. The Central Executive
- Executive control processes govern the sequence of your thoughts & actions, allow you to set goals
& make plans to reach them, & help you rise above habit/routine. However, they can only work on one
task at a time (limits on multitasking).
- The same processes control the selection of ideas active at any moment, & these active ideas
constitute the working memory. Therefore, we link executive control with working memory.
- Emphasize that working memory is fragile; each shift in attention brings new information into
working memory, displacing earlier items.
Entering Long-Term Storage: The Need for Engagement
Two Types of Rehearsal
- Rehearsal generally just means “thinking about”
- Maintenance rehearsal: simply focus on to-be-remembered items, with little thought about how what
the items mean or how they are related. This is a mechanical process; recycling items in working
memory by rep