Lecture 17 - Longterm Memory II Flashcards
What are the factors that affect how we successfully transfer information from short-term/working memory to more “permanent” long-term memory?
• How we encode in STM/WM determines what information is
passed on to long-term memory.
• Encoding can be influenced by five main factors: rehearsal time, levels of processing, memory for meaning, organization, and elaboration.
Once you have encoded information, it is only useful if
you can retrieve it later.
People may use a retrieval plan to
organize the manner in
which they recall information.
trying to match the context at the time of retrieval with the context at the time of encoding
the original format is used to try and retrieve the info
In general, you want the retrieval context to match the
encoding context for maximal efficiency (e.g. the cue should
match the elaboration.)
the cue should match
the elaboration
Thomson & Tulving (1970)
encoding specificity effect.
how important is matching that context at retrieval and encoding?
examined how encoding specificity could overcome pre-existing associations. [Different from example in book.]
• Phase I presented word pairs that were either weakly or strongly existing associations. The first word is a cue, the second is the target.
e.g. =
Weak: ground-COLD,
grasp-BABY
Strong: sky-BLUE,
long-SHORT
• Phase II presented a pair of words and asked if the UPPER CASE word
was previously shown (recognition task). However, they varied the cues
to be either weak or strong.
Weak cue: ground-COLD, pretty-BLUE
Strong cue: long-SHORT, infant-BABY
Which would be better during recognition: strength of pre-existing association or specificity of encoding context (the initial specific encoding cue)?
The amazing find was that a weak retrieval cue was more effective than a strong retrieval cue if it matched the exact cue used when studying.
• You encode the target word with that specific cue.
• Stronger cues aren’t as
effective if not present at
encoding.
• The lure (new words that had never been there) was never present in phase I. You should always say “no” during
recognition.
Context overcomes preexisting associations.
Context overcomes
preexisting associations.
context effects
- you encode the target with specific cues
- are a form of encoding specificity
• The context includes all factors/cues that might be encoded along with the target item.
- it’s your entire environment
• No two environments are identical, but can be similar.
Godden & Baddeley (1975)
context effects
SCUBA
• had scuba divers learn a list of words either on land or 20 feet under water.
- (very different contexts!)
- Participants were then tested either on land or under water.
- Recall (subjects being able to produce the word) was almost 50% better when the encoding and testing occurred in the same setting.
Retrieval: context effects
How might context effects be useful when studying for an exam?
Study in the room where the exam will be given.
state-dependent learning
Your internal state is part of the encoding context. (the context that is internal)
In general, it is found
that retrieval is best when your current internal state matches your internal state during encoding.
Retrieval: state-dependent learning
Eich & Metcalf (1989)
examined the effect of mood on encoding and retrieval.
- Participants listened to happy music and thought pleasant thoughts or listened to sad music and thought depressing thoughts.
- Once the mood was established, they studied a list of words.
- They returned two days later and followed the same mood procedure.
- Results: people who were originally sad were much better if tested in a sad state and went down if they were tested in a happy state (and vice versa)
idea was that just like the external context, if you have those some emotional cues you’ll be better if the internal encoding context matches the retrieval encoding context
Retrieval: encoding specificity
Which of the following exam formats would (generally
speaking) take most advantage of encoding specificity?
true/false
would match the closest
When you practice new (and old) material, you have a chance to re-encode
the information.
can add cues, depth, elaboration to re-encode
Similar to rehearsal in STM/WM, this practice will improve recall.
Retrieval practice
• Roediger & Karpicke (2006)
compared two types of retrieval practice:
re-reading information vs. testing.
advantage in one or the other?
- Participants read a passage and either re-read the passage or took a recall test.
- Both groups then took a recall test after a variable delay period.
- Called the testing effect, it was found that testing improved subsequent recall relative to rereading.
Results: testing was better in the long term
testing effect
if you’re studying and you’re practice style is to test yourself: that’s going to have a better retrieval outcome in the long-run
b/c you’re self-generating: adds depth, elaboration
The manner in which you ___ and _____ also adds
information to the memory.
practice
encode information
(a way of elaborating)
MacLeod et al. (2010)
Production Effect
• presented lists of words to subjects. Half of the words were read silently, while the other were read aloud.
• A recognition test was given after the entire list was presented.
- was this present before or not?
• Results: There was a small, but reliable, advantage for remembering words that had been spoken aloud.
According to this production effect,
the words read aloud are more likely to be remembered because the
additional processing makes the words distinctive (compared to those
read silently).
you make something special about the words read out-loud
multi-sensory encoding
key thing that only SOME words were said out-loud NOT all of them
Re-encoding after retrieval also allows you to encode a new context.
It is thought that distributed practice…
….can be more efficient
than massed practice.
Spacing Effects
In one study, Muller & Pilzecker
- presented two lists of nonsense words to participants. The second list could be presented immediately or 6 minutes after the first.
- They found that the delay group had much better recall for the first list.
spacing effect
distributing your practice is more effective because you allow the info to be fully encoded
suggests that distributing practice of items gives more time for the items to be encoded and consolidated in long-term memory
In general, it is found that spacing study time results in better retrieval, especially over longer periods of time. Studying the same material in different contexts may add more retrieval cues and make the memories more resistant to encoding
specificity effects.
When we forget, is the item
permanently lost or is it still in memory?
• This is difficult to assess, but it seems that forgetting is really a loss of appropriate retrieval cues.
you lose the ability to get a cue that’s effective in pulling these memories out of LTM
In intentional forgetting
experiments,
subjects are given multiple lists to remember. Some lists contain a signal to forget that list
• The ‘forgotten’ items are often not recalled during testing, but may be
recognized in a new list.
• This suggests the memories still exist, but require better retrieval
cues.
What would cause us to forget something in long-term memory?
3 different theries:
decay
interference
overwriting
decay
According to the early “law of disuse,” memories
that aren’t rehearsed lose strength over time.
- idea of memory being like a memory trace: the memory decays: if you don’t rehearse those memories the neural connection “unwinds” or goes away
if neurons stop firing together the neural connection goes away
interference
Perhaps earlier items are more difficult to remember because new material interferes with retrieval?
the cues were the same and now the more recent cue points to a different memory
overwriting
maybe new memories can replace older (similar) memories?
based on computer metaphor
you have neural storage and new memories are encoded over older ones
The memory decay hypothesis is difficult to support
• Anecdotally, we can remember things from long ago if only give the right
cue (e.g. Who was your best friend in kindergarten? When was the last time you thought about them?)
• If time without rehearsal is the key factor, this should be experimentally demonstrable.
Jenkins & Dallenbach (1924)
decay testing
− presented lists of items to be remembered.
− Participants then either slept or resumed normal activities.
− It was found that those who had slept could recall more items.
- Therefore, time can’t be the (only) critical factor.
Interference effects have been shown to occur in a number contexts.
in retroactive interference
in proactive interference
In retroactive interference
there is difficulty in recalling earlier material due to exposure to new material later. (people learned a list, did new different things)
− Learn list A, then List B (or rest).
− Worse memory for items on list A for those who studied List B.
In proactive interference
early learning impedes the memory of new material.
− Learn List A (or nothing), then learn List B.
− Worse memory for items on List B if you had first learned List A.
first list interferes with the second list
sleep encoding
during sleep you’re not interfering with stuff and then the brain rehearses stuff
encoding is a physiological process
