L12 - Metacognition and Eyewitness Recall Flashcards
Is it possible to boost an eye-witnesses memory?
No
They either see things or they don’t, trying to improve their memory is more likely to influence it in a way that makes it less true.
What are the two things that “guide” a memory report in the model of monitoring and memory control
Monitoring of memory (when people are trying to recall)
The things we can tweak to produce differences in output
What did Fisher and Geiselman (1992) find when they looked at how police interviews were being done at the time?
They were based on a lot of folklore and there was very little empiricle evidence guiding how they were being conducted.
What interview style did Fisher and Geiselman (1992) develop?
Cognitive Interview
What are the 4 stages of Fisher and Geiselman’s (1992) cognitive interview?
- Build rapport (foster control over report in the witness)
- Context reinstatement (“think back to the event”)
- Open ended report (free recall)
- Focussed report (re-examine report from difference perspectives etc)
In Fisher and Geiselman (1992) cognitive interview, we are assuming that the witness is….
Co-operative
Not trying to decieve
What does “foster control over report in the witness” mean (stage 1 of cognitive interview)
Handing control of the report over to the witness
What stage of the cognitive interview may this question be a part of - If a policeman asks a witness to “imagine yourself from another perspective”
Stage 4
What % increase of correct detail does the meta-analysis show that the cognitive interview generate?
34%
(some research to suggest increase in false details too)
As there seems to be some research to suggest that the cognitive interview increases false details, how might we address this?
Reducing report threshold
(signal detection theory - report threshold is too wide)
What have studies revealed in regards to which is the least effective stage within the cognitive interview?
The focussed report stage
Which stage of the cognitive interview elicits the most useful and accurate info?
Free recall stage (Roberts & Higham, 2002)
False info can come in when interviewers begin asking leading questions
The cognitive interview is the gold standard of interviews, but what are some problems with the cognitive interview (CI)?
- Relative lack of ease of use for ‘front line’ officers
- Therefore, few police officers use it
- No clear picture of factors driving a improvement in recall accuracy (with police mediated review in particular)
People are setting a report threshold somewhere, but a good memory report is reliant on good memory monitoring
How do witnesses monitor their memory accuracy?
- Meta-cognitive feelings
- Meta-cognitive judgments
What types of monitoring information (cues) that can influence memory reports do people use to determine whether a memory is correct?
Memory content (analytic cues)
Highly detailed memories are more likely to be correct
Perceptual Fluency (inferential cue)
Busey, Tunnicliff, Loftus & Loftus (2000)
Rerieval Fluency (inferential cue)
Repeated post-event questioning inflates confidence (Shaw & Mclure, 1996)
What can happen if you are interviewed by multiple people about what you witnessed?
You generate retrieval fluency which increase your confidence of your account but not accuracy.
Why is it important to allow the witness to say “I don’t know” when asked about whether they remember an event
Control over the report options can influence the answers they give. Witnesses must volunteer the information they remember.
If witnesses are forced to answer, they may give incorrect information
What is “grain size”
What are the two types of grain size?
They degree to which you can describe what you saw
Fine Grain vs Broad Grain
Fine detail vs general information
Why is allowing witnesses control over grain size important in an interview?
It allows them to answer a question without having to say “I don’t know” but give less informative answers
Gives the witness control over report (first stage of the cognitive interview)
What was the two-phase methodology of the Weber and Brewer (2008) experiment?
- Asking someone to generate a reponse to a question and to give their confidence in their answer (but not give it)
- They come back later and decide whether they want to volunteer info
* Ask people to generate answer in fine grain or coarse grain of detail*
What happens to the grain size if you allow people to control over grain size?
What were the results of the Weber and Brewer (2008) experiment?
You can get an increase in the amount of fine-grain responses at phase 2 when you allow people to vary the detail of the information given
Whats the relationship between grain size and accuracy?
What conclusions were found in the Weber and Brewer (2008) experiment?
Giving witnesses control of grain size and you can boost accuracy
If you force them to give a particular type of grain size then you get a reduction of accuracy
What two types of “control” is best for generating the most accurate information from witnesses?
Allowing witnesses control over report size and grain size
Allowing them to say “I dont know” and to choose whether they want to give a fine or coarse grain response
What is meant by the fact that an interview is “dynamic”?
Interviewers can be influenced by inverviewees and feedback occurs in both directions during an interview
In what ways can subtle and explicit cues influence an interview of a witness? (Douglass, Brewer & Semmler, 2009)
Subtle and explicit cues from an interviewer can totally change estimates of view, attention, confidence
Subtle and explicit cues about accuracy in an interview could also affect control of the eyewitnesses report
What is the post identification feedback effect? (sometimes just called the feedback effect)
Confirming and disconfirming information about the quality of the witnesses memory can modify the reports about viewing conditions
In the Douglass, Brewer and Semmler (2009) experiment, what predictions did they make regarding positive feedback?
That positive feedback and an inference that it reflects good memory would lead to a reduction in the output of quantity but would increase overall accuracy.
Why did Douglass, Brewer & Semmler (2009) predict that giving positive feedback may reduce witnesses report threshold?
Producing info and getting feedback is “good” may make them require more confidence to report new information
What did Douglass, Brewer & Semmler (2009) experiment show regarding matching the goals of interviewers and interviewees?
When goals of interview and interviewee match more accurate information is elicited.
- Making sure they have a clear understanding of what the outcomes they want are is a very important step before conducting the interview*
- (matching goals on the right hand side, non-matching goals on the left (not labelled correctly)*
What did Fontaine & Semmler (2016) investigate in their experiment?
- Does mental reinstatement of context impact narrative recall (grain-size, quantity and accuracy)
- How does the instruction to report detail impact on narrative recall
- Do naivety instructions impact narrative recall
- What does the “do not guess” instruction do to narrative recall
What are “naivety instructions”
When you tell the witness “I haven’t seen this, I wasn’t there I can’t confirm it, only you can tell me what went on”
What were the 4 instructional demands given to witnesses in the Fontaine and Semmler (2016) 3 experiments?
Mental-reinstatement-of-context (MRC)
Do-not-guess (DNG
Report-detail (RDT)
Naivety (NVT
What did Fontaine and Semmler (2016) find when giving witnesses the report-detail instruction?
Report detail was a changing criterion
Got an increase in correct details, an increase in incorrect details (errors) and an increase in the fine grain details (precision)
but accuracy was not effected
What did Fontaine and Semmler (2016) find with the mental-reinstatement-of-context instruction?
Iincrease in correct answers - moderate increase in errors but no change in grain size (precision) or accuracy
What did Fontaine and Semmler (2016) find with the naivety instruction?
Moderate increase in correct info
No effect on errors
Large increase in precision
No affect on accuracy
Telling witnesses you weren’t there and you are naive makes them much more likely to give fine grain responses
What did Fontaine and Semmler (2016) find with the do-not-guess instruction?
Slight increase in correct info
No affect on errors, precision or accuracy
Why is it important to have “blind coders” for doing eyewitness research?
It is highly subjective and therefore this increases inter-rater reliability
