Week 5 reading- What Do We Know About Eyewitness Identification? Flashcards
What are some proposed solutions to the eyewitness identification problem?
-Use of Sequential Lineups
-Blank Lineups
-High Functional Size
- High Propitious Heterogeneity
Sincerity
Eye witnesses that make mistakes in Identification do not do so with intent to deceive. They are genuinely mistaken and so fairly persuasive
in testimonies because they believe fully in what they saw
What are the three observations that the eyewitness identification problem is based on?
-Eyewitnesses in psychology experiments often make inaccurate statements about a previously witnessed event
-Their false statements appear sincere
-Faulty eyewitness accounts have been implicated as the primary single cause of documented false convictions
In studying the eye witness identification problem what techniques have been used to present simulated events?
-Slide Sequences
-Film Scenarios
-Live Stagings
Typically the subject is unaware that they will be tested for their ability to remember identify a perpetrator later on.
What is a problem with using real data from actual cases to study the eye witness identification problem?
Superstitions must be made about what actually occurred because the experimenter did not control the situation. It is hard to tell therefore in real cases whether the witness was correct in their identification. Even multiple witnesses in consensus can not be deemed as conclusive.
What is the difference between system variables and estimator variables?
-System variable: Critical variable that which are manipulated in the experiment that can also be controlled in actual an case e.g. line-up construction. These are known as preventable errors and as such previous research has a bias towards studying them because the information gained could be applied practically to better the legal system.
-Estimator variable: sources of eyewitness error or accuracy that are beyond the control of the criminal justice system. For example, the amount of time the eyewitness had the culprit in view. While this could be manipulated in an experiment it is not something we can control in an actual crime situation therefore it is an estimator variable.
People have argued that witnesses in an experiment behave differently to those in a real life crime situation and therefore, results from studies on the eye witness identification problem cannot be generalized to real life situations? Is this justified?
-The surprise element of experimental studies in this area (i.e. that participants do not know they will be asked to recall/ identify a perpetrator) mitigates this issue to some extent as it means that the experiment is mimicking real life conditions.
-Also the fact that participants in most cases believe the crime is real and continue to do so throughout the identification process.
“If subjects believe they are witnessing a real crime, for instance, in what important ways are they different from people who witness a real crime”
-Comment on the participants generally used in this type of study (college students)- although it is true that very young children and elderly tend to be less accurate than the typically college student subject, there do not appear to be any fundamental differences in the principles governing their performance as eyewitnesses
What are the two functions of a line up?
-Establishing proof about the identity of a suspect
-Establishing that a suspect and culprit are not the same person
What are the two general approaches to describing the uncertainty-reduction function of lineups?
-The outcome of the lineup task is an event that increases or decreases the subjective probability that the suspect is the culprit relative to the probability before the lineup
-Alternatively, Navon (1990) proposes that the outcome of a lineup task informs us about the resemblance between the suspect and culprit. This occurs even when the witness makes a mistake (i.e. false alarm) as we have learned that the culprit has similar characteristics to the identified person. So all identification behaviours serve to reduce uncertainty even if they are false identifications.
What is the difference between a recall task and a recognition task in relation to the eye-witness identification problem? How do these two factors interplay?
-Verbal description = recall task
-Identification= recognition task
Recall/ description is likely to indicate identification ability. For instance, a description can be highly diagnostic (providing lots of key unique details about a perpetrator) or it can be less diagnostic where a witness says highly generic statements that are unlikely to help with identification amongst distractor foils.
Note: lectures kind of work against this e.g. verbal overshadowing effect and no reliable studies that show highly detailed accounts in verbal description predict accuracy in an identification task (police tend to overstate this effect)
Difference between a target absent and target present lineup?
-Target absent= suspect in lineup is not the perpetrator
-Target present= suspect in lineup is the perpetrator
Note: in real life settings police do not know which type of line up they have constructed
What is a signal suspect model lineup? In what way is it good?
-There is only one suspect in the lineup and the remaining lineup members are known to be innocent (foils)
-Good because if the witness identifies a distractor/ foil as the suspect in a single-suspect model this would be dismissed as error because we know the foils are innocent. -Therefore, this version of the lineup is deemed as ‘safer’ because it mitigates some error.
-This is the standard way lineups are constructed, at least in NZ!
What is an all-suspects lineup? What is the consequence of using this method?
-All suspects lineup is where there is no distractors. Instead, each member of the lineup is suspected by the lineup operators of being the perpetrator of the crime.
-Using an all-suspects line up increases the chance of error (false-identification) as there is no opportunity to reveal eyewitness error through knowing that the
foils are innocent
In what ways is a lineup similar to a normal psychological experiment and what does this mean for how we construct them?
-A line up is essentially an experiment (has a hypothesis which is being tested) and is prone to errors that can be corrected via construction in the same way.
-One problem that cannot be fixed as it would
be in a psychological experiment is sample size.
This cannot just be boosted to increase certainty in identification as there is a fixed number of witnesses for any given crime.
Blank line up control
A blank line-up contains no suspect, all members are known-innocent distractors. It’s like a noise-alone trial in a signal detection task. Those who don’t identify someone in this ‘fake’ line-up are usually far more reliable when they do make an identification in subsequent line ups.
Mock Witness Control
Compare the actual witness’ response to that of mock Witnesses. If mock witnesses tend to choose the suspect based on limited information provided about the case it points to the fact that the actual witness’s identification was based on the same limited Information= can help clarify how a witness came to make the decision they did
Do we tend to process faces ‘piecewise’ or holistically? Is this for children as well?
-Adults tend to process faces holistically i.e. features are ‘computed’ in relation to each other not as separate distinct entities (distance between the eyes for an Example)
-Children tend to process faces in a more feature by feature manner interestingly (only shown in limited studies though)
What is thae difference between parallel processing and serial processing of faces?
Parallel processing method for faces: sameness judgements are done via this method and do not vary according to the number of features that differ
between the two faces (latency does not differ)
Serial processing method for faces: Difference judgements are done via this method and it takes longer for an individual to process the difference between two faces if they only differ by 1-2 features (difference in latency)
Some researchers propose the methods work in conjunction i.e. we process outer features of faces in parallel and inner features serially.
Importantly, both can be used to describe how we form a holistic view of the face
How does recognition of faces differ according to distinctiveness? What method of presentation has this led to?
-Faces that are atypical are more easily recognised than typical faces
-This lead to people thinking that caricatures that accentuate the unique features of person could be used to help mitigate the eye witness identification problem.
Evidence: people’s faces were more easily recognised with this method than when line drawings or anti-caricatures. But, studies fail to find where the drawings are superior to photographs
What is prosopagnosia? What can tell it us about how we identify faces?
-Prosopagnosia= specific inability to recognize faces.
-individuals with Prosopagnosia tend to not be able to recognize things other than faces (e.g. different birds)
-It’s likely to more be a general deficit in the ability to make complex discriminations from memory
-This suggests perhaps that we do not have a specific system soley dedicated to identifying faces as researchers have suggested in the past.
What is a way the normal facial recognition tests differ to the task of eyewitnesses?
One way that normal facial recognition tests differ from the task eyewitnesses have to do is that eyewitnesses don’t nor have they ever known the culprit.
What is portrait recognition? How is it different to person identification?
Portrait recognition= this is the task used if you were test recall for those in your old class via their
yearbook pictures (this is what the article says but idk if I agree cause you would have ‘real-life memory’ for them not just in a photo). The same Photo is tested as what was Encoded (stimulus stability is high)
- You are recognizing the portrait/ person as opposed to the actual person
Person identification= recognizing someone who was viewed initially under different conditions. This is the case in a line-up (the stimulus stability is low as there are many things that can be changed about
appearance from the time the individual committed the crime to when they appear in the line-up)
Relative judgements
-Relative judgement process is where an eyewitness chooses the line-up member who most resembles the culprit relative to the other members in the line-up
-This is problematic for target absent line ups
Does telling a witness that the culprit may or may not be in the lineup help reduce relative judgments?
-It’s important to tell witnesses that the culprit may or may not be in the line-up as this reduces the chance that relative judgments will cause the witness to identify a ‘perpetrator’ in a target absent condition.
-However, people base decisions on inference and are prone to conformity, obedience and compliance pressures so simply telling an individual that the suspect may not be present in the line-up is not always going to deter false identifications. This is especially true where ambiguity and authority are prominent.
