Making a Case - Interviewing witnesses
- Interviewing witnesses many give vital evidence when trying to solve a crime.
- Loftus & Palmer (1974) - Demonstrated how leading questions affected the accuracy of recall.
--> memory is reconstructive!
- Can be influenced by cues, schemas or emotions.
This topic involves investigating ways to encourage accuracy and factors/influences that affect the testimony of witnesses:
- Recognising face (Frowd)
- Weapon focus (Loftus)
- The Cognitive interview (Fisher)
Frowd's study - Recognising Faces
Aim: To investigate the relative recognisability of the internal and external features of a face - using a lab experiment.
Experiment 1: 15 males, 15 females - Staff & students (1:30 ratio) all from Stirling University. They were paid £2 to take part. Their mean age was 29.2yrs.
Experiment 2: 48 undergraduates from Stirling University - 21 males, 27 females - ALL volunteers.
- 10 celebrity photographs
- 40 composites produced by E-fit, PRO-fit and EVO-fit (facial composite programs used by the police).
- Faces were cleanly shaven and had no specs on.
- 3 groups/conditions - complete composites, composites with internal features only and composites with external features only.
- Pps. were tested individually on one of the 3 conditions (independant measures design)
- They had to match each composite to celebrity face in their own time
Frowd's study - Recognising Faces (continued...)
Procedure: Exp. 2: Photo Array --> used distracter faces/foils to make task more difficult (faces/foils were either made easy or hard).
- Composites were presented one at a time along with the photo array.
- Pps had to pick out the celebrity face from the array which matched the composite.
- Composites were either of internal or external features.
- External features were sorted 35% correct.
- Internal features were only 19.5% correct
- External feature composites were identified more easily (42%)
- Internal feaures only 24%
Frowd - Recognising faces
- It was a lab experiment therefore task is highly artificial and lacks mundane realism - in real life faces constantly move and change. This may have an impact in the way we remember/recognise faces.
- A standarised procedure is used - easily replicated to ensure findings are valid and consistant/reliable. High reliablility means sound conclusions can be made.
- Ethnocentric - All participants were from Stirling Uni/same area. Results cannot be generalised to other countries/areas/people. - although you could argue that Universities attract many people from different areas and even countries so you have a wider 'range' of people.
- Low ecological validity - artificial task. Findings can't be generalised to everyday life.
- Participant variables/characteristics - Some people may be better at remembering/recognising faces than others. Perhaps even a gender difference?
Loftus's study - 'Weapon focus' effect
Aim: To provide support for the 'weapon focus' effect when witnessing a crime - using a lab experiment.
Participants: 36 students from Washington University, aged 18-31yrs. Half recruited from an ad. and paid $3.50 to participate. The rest participated in exchange for extra credit in their psychology classes.
Participants are told that the study is about proactive interference.
Prodcedure: Two sets of 35mm slides were shown. The 18 slides in each series showed people queuing in a 'Taco Time' restaurant.
- Each slide was shown for 1.5 seconds
- In the control group, the slides showed a person handing over a cheque to the cashier.
- In the experimental condition, the person pulls out a gun instead.
- This difference made up 4 slides - the rest of the slides were identical.
Loftus's study - 'Weapon focus' effect (continued...)
Procedure continued: - There were 4 dependant variables:
- Pps were asked to do a 20-item multiple choice questionnaire.
- They were also given a line up of 12 head & shoulder photos in a random sequence, and had to identify person (with cheque or gun).
- They had to rate their confidence of correct identifcation on a scale of 1-6 (1- guess, 6- very sure).
- Eye movements (fixation & duration) was also measured.
- Eye fixation data showed and average of 3.72 on gun and 2.44 on cheque.
- Cheque condition - 38.9% chose correct person
- Gun condition - 11.1% chose the correct person
- Answers to questionnaire about the slide show showed no significant difference.
- Chance performance on photo line-up was calculated to be 8.5%.
Loftus - 'Weapon focus' effect
- Ethnocentric - In USA guns are legal (illegal to conceal weapons) so perhaps people would be more used to seeing them and so focus on them less than perhaps in other countries like the UK.
- Reliable - Standardised procedure which can be easily replicated. High reliability means more 'sound' conclusions can be made. Also Loftus did a second experiment, using the same procedure with another 80 psychology students, and this supported the findings.
- Lacks mundane realism - Task is very low in emotional arousal so perhaps less time was spent looking at the weapon than in the real situation. In 'real life', we spend more time focusing on the weapon because it's anxiety provoking.
Fisher's study - The Cognitive Interview
Aim: To test the cognitive interview (CI) in the field - a field experiment
Participants: 16 serving detectives from Florida. All with minimum of 5yrs in the division.
IV - Standard interview or cognitive interview. DV- Objective statements.
- Detectives were aksed to record a selection of their next interviews using standard techniques they would normally use - this took 4 months and 88 interviews were recorded.
Interviews were mostly related to robberies and bag snatches
- Detectives were then divided into groups: One group was trained in CI techniques (7 detectives) - Four 60 mins training sessions.
- Over next 7 months, more interviews were recorded by the 2 groups.
- These interviews were then analysed by a team at the Uni of California who were blind to the conditions - the amount of info gathered by the 2 groups was collated.
Fisher's study - The Cognitive Interview (continued...)
- 63% more info was obtained by detectives trained in CI.
- Pre & post training = 47% more info
- Validity of statments backed up by secondary source(s).
- Field experiment - high in mundane realism & ecological validity.
- Qualitative data - which was operationalised so easy to analyse and less time consuming.
- Participant variables/differences - Some pps/detectives may be better at conducting CI, or have better rapport, than other detectives so can acquire more info (lacks internal validity - to improve this, increase the sample size).
- Ethnocentric - All detectives from Florida. Results may not be reliable if repeated in different parts of the world and so results cannot be generalised.
Key words & meanings:
- The Cognitive Interview - A set of instructions given by an interviewer to a witness which is designed to facilitate accurate recall of the original event & to search through a memory using a variety of retrieval methods.
- Proactive interference - When something you learned earlier interferes with your memory of the present.
- Internal validity - whether a study has been affected by biases or errors in the study design.
- Facial morphing - taking several 'original' versions of a face and merging them together by using points mapped onto the face electronically.