Standard Model of Memory
o Working memory vs. Long-term
o Sensory Memory
Sense receptors stay activated for a brief duration after they have
stopped receiving stimulation. This lingering activity is referred to as
o Sperling (1960) – How much information is available via iconic memory?
o Participants were briefly flashed a grid of letters, and then had
to report as may letters as they could remember seeing.
o Participants can correctly recall 4.5/12letters on average
Why aren’t people seeing all 12?
o Theory 1:
Iconic memory (visual sensory memory) is degrading and note
o Theory 2:
Iconic memory only lasts for so long, it fades too quickly. o Sperling (1960)
o Participants would be briefly presented with a grid of 12
letters. Then, an auditory response cue would be presented.
Participants would then have to report what they had been
visually presented with.
o Response cues are one of three tones (high, mid, low), each
tone would mean they would have to report a specific row.
o Theory 1: You should only get a couple because much of it will
o Theory 2: People should be able to report a whole row
o We see a combination of both theories
o Participants were able to report 3.3/4 of the letters,
irrespective of which row they are cued to respond for.
o Participants are seeing an average of 82% of the total numbers
presented. However they only report 4.5/12 of the total letters
because iconic memory decays quicker than the letters can be
o Sperling (1960) – How quickly does iconic memory fade?
o Similar to last experiment
o Similar to last experiment, plus additional manipulation: tone
occurs between zero seconds and 1.0 second after the letters
o They remembered less as more time passed
o “Short term memory”
o Holds the knowledge we are currently aware of
Content easily lost if not attended to Association
o Memory is associative:
When one memory is linked to another, the second becomes easier to
recall when the first is brought to mind.
o If we group pieces of knowledge together into a larger unit, each of the
individual pieces will be easier to remember
Short-term Forgetting: Brown-Peterson Paradigm
o Information presented just before a distraction tends to be forgotten easily
(ex: people’s names)
o Murdock (1961)
o Participants were presented with a set of stimuli to remember,
followed by a number. Participants would have to count
backwards from the number for a period of time, and then try
to remember the original stimuli.
o 1) Original stimuli were three consonants, three words, or a
o 2) Participants had to recall the stimuli after a day of 3-18
o Proves short-term forgetting
Support for Trace Delay
o The previous experiments were presented as support for decay in short term
memory; despite not overloading participants’ short-term memory capacity,
knowledge is systematically lost over time.
o Two types of interference are relevant:
o When new information impairs our ability to recall old info
o When old information impairs our ability to acquire new info
Why not Interference?
o Why isn’t the case that counting backwards retroactively interferes with
remembering letters and words?
Reinterpreting Brown-Peterson Results
o Keppel and Underwood (1962) noted that participants demonstrated almost
no forgetting on the first trial. However, performance degraded on
o Perhaps stimuli from early trials are proactively interfering with later trials? Release from Proactive Interference
o Loess (1968)
o Participants performed the Brown-Peterson task, having to
remember word-triplets. Triplets would be all members of a
particular category (e.g., animals)
o Word categories would change (e.g., animals, vegetables, names)
every trial, or every 12 trials.
o Wickens, Born & Allen (1963)
o Participants were presented with lists of target items to recall
o Every few trials, target items would be changed from
consonants to digits.
o Maximal recall of lists occurred on ‘switch’ trials. Performance
degraded on each successive consistent trial. Decay
o If information in working memory is not rehearsed, it is rapidly lost
o Information in long term memory is believed to be more permanent
Serial Position Curve
Things at the beginning are remembered well because they have been
rehearsed and entered long term memory
Things at the end are remembered well because they have not
decayed from short-term memory.
o When words (or a number, or pictures or anything) are presented in a list,
and people have to later recall the list details, they are best at recalling things
at the very beginning and very end of the list. Main Topics:
Standard Memory Model
We typically treat memories as things that are “stored” for later access.
However, most evidence suggests memory is something that is “reconstructed”
rather than “retrieved”
o Passive exposure to information is typically not enough to remember it.
Knowledge has to be actively worked with for reliable remembering.
Level of processing Type of encoding Exame of questions used
to elicit appropriate
Shallow processing Structural encoding: Is the word written in capital
emphasizes the physical letters?
structure of the stimulus
Intermediate Processing Phonemic encoding: EmphasizesDoes the word rhyme with
what a word sounds like weight?
Deep Processing Semantic Encoding: EmphasizesWould the word fit in the
the meaning of verbal input. sentence “He met a ___ on the
o When we are paying attention to something’s descriptive properties, we are
focusing on surface characteristics:
Modern psychology was founded in 1878
The forebrain has 4 areas called…
Reinforcement increases the likelihood of behavior
Punishment reduces the likelihood of behavior
o Trying to remember information based solely on surface characteristics is
challenging and inefficient
o When we are paying attention o how knowledge could be used, or how it
relates to other things, we are paying attention to deep characteristics.
If I want to keep my little brother out of my room, would
reinforcement work better as a tool than punishment?
Titchner’s structuralism and Plato’s theory of forms are similar
because… Surface vs. Deep Characteristics
o We better remember information if we focus on deep characteristics
o Recall: memory is associative
o The more associations you can form between knowledge, the easier all of the
associated bits are to recall.
o We better remember information if we can attach it to a larger context
o Craik & Tulving (1975)
Participants were presented words in various contexts, and
later had to try to recall the words they saw.
Words were either presented in sparse contexts or rich
She cooked the ____
The great eagle swooped down and carried off the
When later tested, words in the rich contexts were better
o Sometimes it is possible to tae surface characteristics and make them deep
characteristics by embedding all of the pieces of knowledge into a greater
whole. Ex: FTOP = Frontal, temporal, occipital, parietal
o Presenting information in a visually plausible way, in addition to just as
words, makes that information easier to remember.
o Wollen, Weber and Lowry (1972)
Participants were presented with two pieces of information,
and had to recall that information at a later time. A visual
picture depicting that information was presented alongside the
word to express it.
Pictures were of two things, either interacting or not
interacting AND depicted bizarrely or normally.
Participants best recalled information when it was presented
as interacting, and when it was presented in a non-bizarre
fashion. Encoding Specificity
o States that we encode information along with its context. Retrieval cues
should be more effective if they were present during encoding.
o Godden & Baddeley (1975)
o Participants are presented with a list of words to remember.
They are later given a free recall test.
o 1) Studying either occurred on land, or underwater and 2)
Testing either occurred on land, or under water
o Grant et al (1998)
o Participants were given an academic journal article to read.
They were later given a series of short-answer comprehension
questions to complete.
o 1) While reading, participants wore headphones that either
played ambient chatter from a cafeteria, or 2) were silent.
Testing conditions were similarly manipulated.
o Matching learning context to testing context means that we do
better and remember more.
o It is the idea that internal states (ex: emotion, hunger) also become part of
context at time of memory encoding
o Learning something when we’re hungry – remembering it better when we’re
o Our internal states and disposition also become apart of memory encoding
o Eich & Metcalfe (1989)
o Participants were given a list of words to remember. They then
came back 2 days later and tried to free-recall the words
o Prior to study and testing, participants’ moods were
manipulated to be happy or sad. Manipulation was performed
by having people listen to “merry” music or “melancholic”
music, and attempting to recollect similarly valenced events
from their past for 15 minutes.
Results o Studying the words in a happy state – remember them better
when they’re happy. Same goes for sad.
o Suppose you have 8 hours to study for a midterm
o You can spend all of this time studying the night before, or you can spread it
out over a week
o How should you best study for your midterm? How do you allocate that time?
o Spreading studying out results in better long-term retention of knowledge.
o Standard Memory Model
o Storing Knowledge
o Emotional Events
We typically treat memories as things that