lec 4 - reconstruction and false memory Flashcards
what happens over time when you store a memory
regularisation, schematization, reconstrcution etc
-lose the finer grain detail, high levele episodic detail
-begin to similify over time
reconstruction of images exp
-carmicheal et al 1932
Participants were shown pictures
with verbal label and asked to draw these pictures later on.
eg same picture with different labels
‘curtains in a window’
or
‘diamond in a square’
Results: drawn pictures “matched” the label.
what are the advantages and disadvantages of labels activating schemas
Advantages:
We do not need to store exact copies of the information we process.
→ Schemas facilitate encoding, simplify storage, and guide retrieval. In other words, schemas guide/shape online processing and memory representations
Disadvantages
Schemas do not preserve episodic details.
Thus when recalling information – particularly at a delay – we might generate or construct a reasonable, but potentially distorted, memory
explain semantic structures –> semantic intrusions
We remember the gist; we forget the details
* remember the verbal label, forget the perceptual form
* remember the meaning, forget verbatim information
(memory distortions, false memories)
semantic intrusions are false memories
Paradigm 1 ; DRM
Roediger and McDermott (1995)
-drm experiment gives you a list of words to study
-and then you get a test
either recall (write down what you remember) or recognition test (given words some are in list and some are not and you tick off the ones that are)
-
Recall of studied items: 62% (Exp2)
-normally find serial position curve (items begginning are remembered well and items at end of list also remembered well)
-people added items that were never given
the words were about sleep and that was one of the falsley recalled word ‘sleep’ 55%
-
Recognition:
Studied items: 57% “remember” responses, 22% “know” responses
Critical items like ‘sleep’ (never on the list : 58% “remember” responses, 23% “know” responses
Subsequent research: false memories are less vivid
explain how false memories are ‘logical’ errors
We “fill in the gaps” with information from LTM
what is fuzzy trace theory
(Brainerd & Reyna, 1990)
when you encode info,you store two types of traces
gist- semantic meaning
verbatim - ability to store low level detail, visual details etc
Learners store a gist trace and a verbatim trace.
Verbatim traces are weaker than gist traces.
Learners generate false memories if they rely on gist traces
what is the activating/ monitoring framework
false memories are generated by:
* activating related information in LTM
* and then failing to reject this information as not having been studied
activation and monitoring framework
‘activation’
Activated information activates related information, either consciously or unconsciously (Roediger et al., 2001)
-Spreading activation in
semantic networks
activation and monitoring framework
-contributing factors
More associations → stronger activation → more false memories
Strength of associations (eg strongly related words)
Number of semantic associate (how many) more likely to get a false memory if there is many associates
Focusing on semantic relationships (if you integrate what your hearing, be more likley to activate the studied items)
what is source monitoring
ability to recall or judge the origin or source of a memory (Johnson et al, 1993).
- Did you see event X on TV or did you read about it online?
- Did you hear about event X from one friend or from another friend?
what is source misattribution
If we don’t remember the original source, we may misattribute a memory to the wrong source
→ source misattribution (Lindsay & Johnson, 1989)
reality monitoring
Reality monitoring:
ability to discriminate between internally-generated and externally-generated memories
(Johnson & Raye, 1981).
Did you hear the word “sleep” during the study phase of the experiment or did you activate the word “sleep” yourself?
monitoring failure
-exp (example)
If asked explicitly, participants are unable to correctly identify the source of words in DRM lists (Payne et al., 1996)
-* Study: watch a videotape showing two speakers (either Jason or Carol) reading out DRM lists
- Test: first complete a free recall test 3 times, then (after Test 3) indicate the source of each word: Jason, Carol, I don’t know (“neither”)
-if we look at recall of items that were studied, people remmeber 30% over 3 tests (not very man)
-critical words eg sleep, falsely recalled at similar levels 30%
-people were very willing to indicate the ‘source’ of words that they had not heard
Poor monitoring → difficulty rejecting intrusions → more false memories
Improving monitoring → lower rates of false memories
distinctiveness
Warnings
Processing time at test
distinctiveness : anytime you encode information, if you can encode distinctive details about that word/picture/ youll be less likely to generate false memories, since distinctive items stand out
contributing factor: monitoring
Distinctiveness
anytime you encode information, if you can encode distinctive details about that word/picture/ youll be less likely to generate false memories, since distinctive items stand out
contributing factor: monitoring
warning
-if we tell partcipants that they will take part in an exp where we know participants will make many errors , that will trigger more careful monitoring, youll engage more , help reduce false memory
contributing factor: monitoring
processing time at least
-because we know monitoring takes time
-manipulating the time we have will modulate false memory rates
-if we give you less time in the test, you wont be able to engage in the test as well and have more error
paradigm 2: misinformation
-introduces misinfo at some point during the test
-and then we test whether that misinformation makes its way into your recall
paradigm works like this
-give you something to study, pictures, words etc
-eventually ask you questions about what you study
-in between this we introduce post ecent information, can take many forms
-eg sometimes a task (ask questions such as introducing a detail thats not in original event)
-we check whether you will remmebr the misinfod detail or the real one
-and then we ask you a question regarding that detail (test of misinformation)
misinformation
planting new memories
loftus and pickrell 1995 paradigm
Can you “recover” memories you thought you didn’t have?
Loftus and Pickrell (1995) paradigm:
- Collect childhood experiences from the parents of young adults.
- Interview the adults repeatedly about memory for these events plus one that had never happened.
- At first, people do not “remember” the new event. But in subsequent interviews, some do report it as a real memory
examples of planting new memories
experiment
eg
Getting lost in a mall
…when you were five, you got lost in a mall where you went shopping with your mom (and siblings…). Somehow you lost your way. You were crying when an elderly lady found you.
Spilling a punch bowl
…when you were five, at a wedding reception, you bumped into a table and spilled the punch bowl on the parents of the bride…
20-40% of participants report memory for these events:
“It was an outdoor wedding and I think we were running around and knocked something over like the punch bowl
or something and made a big mess and of course got yelled at for it”
’ a picture if worth a thousand lies’
study
-plant new memories exp
-took pictures of participants eg child with his dad
-using photoshpp inserted these pictures with other pictures
eg telling the child, when you were five you went on this hot air balloon ride (even though he didnt’)
-and ask them qs about other events that did happen
-then ask them over several interviews over several days about the story of what happened that day
-
’ a picture if worth a thousand lies’
study
-plant new memories exp
RESULTS
-in the first interview , there are already some partcipants who say this false event actually happened and they remember
-by last interview, more people remember these false memories
reconstruction of stories
What happens when there is no schema? (you dont have information to rely on as a crutch to help you fill in the gaps)
story reconstruction
Bartlett’s “The war of the ghosts” retellings (1932) Participants (Edwardian England) read a Native American folktale, then recall it multiple times at longer and longer intervals.
-asked to read a fairytale (old cultural , not familiar)
-they were asked to read it a couple times , and then they were invited back into the lab at various intervals, and they were asked to recall the story
Reconstruction of stories
when people Recall content:
evidence of forgetting:
retellings got shorter over time
multiple omissions and inaccuracies
no supernatural elements (first thing to go, hard to incorporate into gist memory)
Recall errors:
evidence of distortion and reconstruction:
retellings are more coherent, more “rational” substitutions (“canoes” recalled as “boats”) but schema-consistent information remains intact
reconstruction of stories
Bergman and Roediger (1999) replication
-modern day replication of the reconstruction of stories study,
-new participants also unfamiliar with the culture this story comes from
Study showed :
Recalls scored for major and minor distortions (e.g., substitutions, inferences, intrusions).
number of accurate details that participants were able to recall was not very high just after 15 minutes reading the story
if we bring them back to lab a week later or 6 months later, the number of accurate details drop
-
Results show loss of accurate information and proportionately more distortions over time
they had an additional condition: there was a group in this study that didnt get the first test 15 minutes after studying, they came bavk one week later, and came back 6 months later
-the participants who were not given the initial test, compared to who did: showed testing effect, this shows that engaging in a test or retrieval attempt right after encoding, will help you remember that information better
the process of recalling information does not leave….
the process of recalling information does not leave the memory trace for that information intact
summary
-can we easy distinguish true memories from false ones
We can’t easily distinguish “true” memories from “false” memories by examining overt behavior: people behave as if they have in fact experienced the events they are recalling
→ implications for eyewitness testimony: Can we use memories/confidence as evidence?
