For. 2 Making a Case Flashcards
(69 cards)
Bruce - BACKGROUND?
E-fit faces
Delay, our cue dependent memory and individual differences in memory ability affect facial memory
Composite photos have a limited number of face shapes/features and assumes that we recognise faces by breaking them down into parts
E-fit is hollistic and has a large database
Bruce - AIM?
E-fit faces
Investigate recognisability of internal and external features in facial recognition
Bruce - SAMPLE (exp.1)?
E-fit faces
30 (15m / 15f)
staff/students from University
Independent measures - 3 conditions
Bruce - SAMPLE (exp.2)?
E-fit faces
48 undergraduates (21m / 27f)
Bruce - PROCEDURE (exp.1)?
E-fit faces
3 conditions: complete composites, internal features, external features
shown 10 celeb photos and 40 composite (e-fit) images
told to match the celebs to their e-fit
Bruce - PROCEDURE (exp.2)?
E-fit faces
2 conditions: easy (all different) and difficult (all similar)
photo array of celeb faces
matched with composites (external or internal) with distractor faces
Bruce - RESULTS (exp.1)?
E-fit faces
35% of complete and external features composites were matched correctly
19.5% of internal features composites were matched correctly
Bruce - RESULTS (exp.2)?
E-fit faces
42% of external features composites were identified correctly
24% of internal features composites were identified correctly
[Participants in internal condition performed above chance level
Participants in complete and external features condition performed equally well]
Bruce - CONCLUSIONS?
E-fit faces
Shows external features are more important for facial recognition
Faces processed holistically
New facial software allows faces to be changed hollistically
Bruce - EVALUATION?
E-fit faces
Exp.2 has larger sample (80) = more reliable
Standardisation = confidence in results
Viewing slides = lacks eco. validity
Lack of stress = more ethical
Useful = implications for eyewitness/court etc.
Reductionist - doesn’t consider ind. differ.
Psy. science - standardised/controls, recording eye movements = scientific
Loftus - BACKGROUND?
Factors influencing identification
Factors that influence witness accuracy:
delay
stress and arousal (eg level of violence)
weapon focus effect
leading questions
line-up instructions
Loftus - AIM?
Factors influencing identification
Support weapon focus effect when witnessing a crime
Loftus - SAMPLE?
Factors influencing identification
36 students from Washington University
Aged 18-31
Half recruited by advertisements (paid $3.50)
Half were psychology students (given extra credit)
Told were in experiment for ‘proactive interference’
Loftus - PROCEDURE?
Factors influencing identification
Shown 18 slides of man queuing at a restaurant
Control: Man gives cheque to cashier
Experimental: person pulls gun on cashier
Each slide: 1.5seconds
Recognition of man tested - shown 12 photos (line-up) and rated how confident they were and given questionnaire
Loftus - RESULTS?
Factors influencing identification
No difference in answers to questionnaire
Cheque condition: 39% correct identification (2.4secs looking at cheque)
Gun condition: 11% correct identification (3.7secs looking at gun)
No difference in confidence
Loftus - CONCLUSIONS?
Factors influencing identification
Participants spent longer looking at gun so had more difficulty recognising supect from line-up as more time was spent looking at weapon
Witnesses can be unreliable
Crime involving a weapon affects a witness’ ability to recognise a suspect’s face
Loftus - EVALUATION?
Factors influencing identification
Big age range = celeb familiarity would differ
Small sample = not reliable
Standardised, lab = reliable
Eco. validity - lab setting but photo array mimics police line-up
Useful - informs police/courts about witness reliability
Psy. science - standardised, meets criteria
Fisher - BACKGROUND?
Cognitive interview
Standard interview produces unreliable results in memory
Cognitive interview stimulates cues (conttext) to maximise retrieval of memory
Cognitive interview:
-context reinstatement
-report everything
-change perspective
-change order
Fisher - AIM?
Cognitive interview
Compare performance of detectives with and without Cognitive Interview training
Fisher - SAMPLE?
Cognitive interview
16 detectives
Florida police
7 - cognitive interview course
9 - untrained controls
Fisher - PROCEDURE?
Cognitive interview
Detectives tape recorded interviews using standard interview techniques
2 groups formed - 7 (originally 10) trained in cognitive interview and 6 rest were untrained controls
Training was 4x 1hour sessions
Re-interviews by 7 trained and 6 controls
Fisher - RESULTS?
Cognitive interview
47% more information recorded after training
6 (of 7) did better after training
63% more information recorded than control
94% accuracy with witness statements
Fisher - EVALUATION?
Cognitive interview
Small sample = unreliable
Detectives experienced in robbery = not generalisable
Ethnocentric - Florida police
Nature - CI techniques can be learnt
Useful - Improves memory retrieval from witnesses; training relatively easy to give
Fisher - CONCLUSIONS?
Cognitive interview
Cognitive interview is effective
Training is relatively easy to provide