Lecture 22 - Reconstructive Memory in the Real World Flashcards
real world memory
real world context outside the lab with more importance on what your remembering and the accuracy
Eyewitness testimony is a type of
episodic memory
misinformation effect
how episodic memory can be affected when you’re integrating incorrect info
flashbulb
things you feel are very accurate and “burned into your memory”
Which of the following best describes the relationship
between schemas and scripts:
Schemas are less structured than scripts and provide general information about a type of event.
contain many slots for values to be entered for a particular type of event
A schema is a….
….a general knowledge structure used for
understanding complex situations.
A schema is general for a ____ of situation, and helps give
______ to an specific event.
type
organizational structure
Schemas are more than a bunch of facts –
– they are structured to include the relationships between facts
scripts
much more detail
specific order of events within a schema
gives you causal (order) links: tells you how to move onto the next step in the script
Last time: Problems with schemas/scripts
Just as we use schemas to understand (encode) events, we
also use them later during recall.
- Source monitoring
- Sometimes we use stereotypes (a schema for a person) to reconstruct a memory.
• This can be very useful – as with all schemas you don’t need to
recall every detail, but can fill-in with defaults.
• However, this can (and does) lead to many errors.
source monitoring
when you try to ID where did that memory come from: is that what happened last time or the time before that? what is the source of the specific memory that has popped into my head?
is the process of determining the origin of some memory or belief.
stereotypes
systematic errors based on previous episodic info
source monitoring error.
When we attribute a remembered item to the wrong source
memory that popped into your head isn’t from last week it’s from 2 weeks ago
Marsh, Cook, & Hicks (2006) examined the
role of gender stereotypes on source
attributions.
• Subjects read statements (e.g. “I like
baseball.” or “I designed a center piece.”), that
were attributed to one of two people: Chris or
Pat.
• Sentences were masculine, feminine, or neutral.
• Subjects were given a puzzle (delay) and later
told the gender of the people (Chris is a
heterosexual male, Pat is a heterosexual
female).
• They then re-read the statements, and asked the source (who had said them).
• People were more accurate when the statements matched the stereotypes for masculine or feminine (schemas). => 83% of masculine statements were correctly attributed to the male name, but only 65% were correctly attributed to the female name.
Reconstructive memory
When recalling an event, we sometimes…
…. add elements that were not originally present.
• The more complex the original event (i.e. the more it taxed short-term or working memory during encoding), the greater
the…
…. difficulty in recalling exact information.
fill them with a default or something plausable
what determines adding not originally present elements?
- The more complex the original event (i.e. the more it taxed short-term or working memory during encoding), the greater the difficulty in recalling exact information.
- If you later use a cue to probe part of a memory, you may rely on default values when you can’t remember precise detail.
- This use of what was likely, given the schemas, is reconstructive memory.
If you later use a cue to probe part of a memory,
you may rely on default values when you can’t remember precise detail.
facts consistent with the schema but not necessarily with the events as they occurred
reconstructive memory
you’re predicting or estimating what’s most likely based on schemas and scripts
reconstructing the memory based on what makes sense
The schema you use to recall an event may act as a set of cues (i.e. the slots to be filled) for your memory. Anderson & Pichert (1978) asked participants to read a story from the
perspective of a burglar or home buyer.
- The story contained many descriptions of rooms and the valuables located in them.
- When asked to recall details from the story, 64% of perspective–relevant details (e.g. burglar) were remembered, compared to 46% that had to do with the other perspective (home buyer).
- However, if asked to then switch perspectives (e.g. from burglar to home buyer), participants remembered 10% more detail.
- they didn’t re-read the story they just thought about it differently!!
- new 10% comes from new perspective
*the schema you already had in your head, you’re able to adapt that and use a whole different set of cues to pull up new info from LTM to reconstruct working memory - what you’re consciously aware of changes based on schema
• Your working memory for the story was reconstructed based on the schema activated.
—–> within that script all those slots have diff cues and you fill those place holders with diff things from LTM
Eyewitness testimony (episodic memory!!) in a criminal case is convincing in a courtroom because it is largely assumed to be accurate.
The greater the confidence of the testimony, the greater the implied accuracy.
- 75,000 – 80,000 suspects are identified by eyewitnesses each year.
- Witnesses are examined for credibility, internal consistency (do they contradict themselves?), and confidence.
Loftus, Burns & Miller (1978) examined how
misinformation may alter eyewitness memory.
- Subjects were shown slides of a car accident. Half saw slide A (with stop sign).
- Later asked, “Did another car pass the red car while it was stopped at the (yield sign/intersection)?”
- Then given a recognition test: “Did you see slide A or slide B originally?”
- If given neutral information (intersection), they chose A 85% of the time (accurate).
- If given misleading information (yield sign), they chose A 38% of the time (wrong).
- they went with what they were told
• This is called the misinformation effect.
the problem with eyewitness testimony
- it is very difficult to determine the individual accuracy of a witness unless there are other witnesses.
- Because memory is reconstructive, eyewitness accounts can be systematically distorted.
!!you can change memory on the fly if you change perspective - it is very easy to distort memory with misleading information
the misinformation effect
effect on memory where you take something you remember and replace it with new misinformation
Three main exaplanations have been suggested for the misinformation effect.
Overwriting
Source confusion
Misinformation acceptance
Overwriting
The new information erases and replaces the original memory.
(This was the Loftus et al. interpretation.)
Source confusion
The new information competes with the original memory during retrieval - both active in LTM and you just go with the most recently brought up thing.
− Both are present in memory, but difficult to choose.
Misinformation acceptance
You don’t notice the difference between
the two episodic memories or recall the sign at all; you just assume the misleading information is correct (or equally valid).
− Why would you be misled when trying to get to the truth? Assume the second thing (most recent) is true.
McCloskey & Zaragoza (1985) tested the misinformation acceptance hypothesis with a modification of the Loftus et al. (1978) experiment.
- Participants SAW slides of a worker stealing $20 and calculator. The items were placed in a toolbox, under a hammer.
- Later READ a narrative saying the stolen items were placed under a tool (neutral control - could refer to the hammer) or a screwdriver (misleading because there was no screwdriver).
- If later asked which was seen “hammer or screwdriver”, you would expect:
* If read “tool” (neutral) => respond “hammer” more often.
* If read “screwdriver” (misleading) => respond “screwdriver” more often.
===> YES that’s what they found
- MODIFICATION: Did the same thing BUT were asked “Did you see hammer or wrench”?
* If misleading information ‘overwrote’ original memory (hammer), then both responses would be equally likely (chance).=> “Hammer” was the more frequent response, indicating it hadn’t been erased or overwritten. => Evidence for misinformation acceptance hypothesis.
Lindsay (1990) tested the effect of awareness on misleading information. If you can discount the misleading information, does it
still have an effect?
acceptance hypothesis suggests people aren’t aware of the difference between the actual info and misleading info.
what happens if people are made explicitly aware? will there still be errors?
- Subjects watched slides of a man stealing a computer.
- They then read a narrative with misleading information.
• 48 hours later, they are given recall questions, but were first warned to
disregard the narrative because it was wrong.
• Under the misinformation acceptance hypothesis, participants should not report anything from the narrative.
==> However, subjects still reported items from the narrative. This is consistent with the source confusion hypothesis (you come up with a memory and can't tell where it came from: slides or narrative?). - reconstructive nature of memory, whatever is cued the best is what you recall
So which is it? misinformation acceptance hypothesis or source confusion hypothesis?
not clear
- While the basic effect (the misinformation effect) is not questioned, there isn’t an accepted explanation.
- Overwriting is less likely, as many experiments demonstrate access to the original memory when using the right cues.
- Source confusion and misinformation acceptance (if the misinformation is plausible) are likely both parts of the explanation.==> Note that they both take advantage of schemas and scripts = schemas act as a cue