Making a case!


Making A Case

Interviewing Wtinesses

  • Face Recognition- Bruce
  • Weapons Effect- Loftus
  • Cognitive Interview- Gieselman & Fisher 
1 of 14

Bruce- Experiment 1

Experiment 1

Aim: Investigate relative probability of internal and external features of composite faces; the predicition is that internal ones will be of poorer quality than external ones

Procedure: 30 staff and students from Stirling Univeristy; Target photos of celebrities used along with 40 composites of same celebrity; Created by people unfamiliar to faces using sketch art, EvoFIT or e-fit systems

  • Group 1 = complete composites
  • Group 2 = Internal features composites 
  • Group 3 = External features composites

Results: Whole composites and external features were matched best (33%); Internal (19.5); Supports idea that external features were better at created in the composites than internal ones because unfamiliar processing being used

2 of 14

Bruce- Experiment 2

Experiment 2

Aim: Find whether internal features actually offer any useful information

Procedure: 48 students at Striling University ; Group 1= 40 internal feature composites to be sued to select a target celebrity from an array of photos including 'foils'; Group 2= 40 external feature composites

Results: External feature composities (41.6%); Internal (28.4%); Supports experiment 1 and suggests that internal part of the composites was of poor quality

3 of 14

Experiment 3- Bruce

Aim: Find whether problem in creating the internal features is because of the unfamiliarty of the faces or because constructing internal features is just difficult 

Procedure: 32 compostions created; Half faces familiar to editor and half unfamilair; Presented as complete faces OR External featrures OR Internal features

Sample: 54 participants asked to match the composites with the target faces, and to name them if they knew them

Results: Complete faces and external feature faces were matched at the same level; Much better than internal features; No difference between the quality of the features 

Conclusions: Participants preformed just above chanve level for internal features; Participants preformed equally well with whole faces and external features; External features are more important for facial recognition and faces are processed hollistically; Implications for facial reconstructions whcih involve picking internal features from a book; Newer face reconstruction software takes into account E-fit or EvoFIT.

4 of 14

Loftus et al

Aim: Provide support for the weapons focus effect when witnessing a crime 

Procedure: Lab experimentl; 36 students from University of washington; 18-31; Recruitede via advertisement; Paid $3.50 and psychology students given extra credit; Shown 18 slides of a series of events in a taco restaurant; 

  • Control group see a man giving cashier a cheque
  • Experimental group see man pulling a gun
  • DV was recognition of person= measures by 20 item multiple choice questionaire 
  • Patricipants were also shown 12 photos in random seuqence and asked to rate confidence with their identification

Results: Control group gave a 38.9% correct identification; Experimental condition had an 11.1%  correct identification; Eye fixationd ata showed an average fixation time of 3.72 seconds on gun compared to 2.44 seconds in cheque 

Conclusion: Participants spent longer looking at the weapon and therfore had more difficulty in pikcing the suspect from the line up 

5 of 14

Geiselman & Fisher

Aim: To test cognitive interview in the field

Procedure: Field experiment; Interviews with real witnesses by 16 detectives from robbery division of dad county police, Florida; 7 detectives trained in CI; Interviews recorded and analysed by a team at university of california; Blind to conditions

Stages of Cognitive Interview: 

  1. Report everything
  2. Context reinstatement 
  3. Recall in reverse order
  4. Revall from a different perspective

Findings: 63% more information obtained by detectives trained im CI

Conclusion: CI technique do seem to work; More information gathered; Help police solve more crimes

6 of 14

Interviewing Suspects

Interviewing Suspects

  • Detecting Lies- Mann
  • Interrogation- Inbau
  • False Confessions- Gudjohnsson
7 of 14


Aim: Test police ability to distinguish truth and lies during interview with suspects

Procedure: 99 Kent police officers; Participants shown 54 video clips involving 14 suspects of real life police interviews; Clip supported by further evidence which established truth or lie; 6 to 145 second long videos; Particpiants asked to fill out questionaire about experience of detecting liars and decide if each clip was truthful or lie; Asked to make a note of cues they uses to detect lies


  • Polic officers preformed better than chance 66.2% accurates lifes and 63.6% accurate truths
  • Cues = Gaze, Movements, Vagueness, contradictions

Conclusions: Police officers good at detecting lies; Study did not have a control group because of issue of privacy and confidentiality 

8 of 14


  1. Direct confrontation
  2. Chance to shift blame
  3. Never allow suspect to deny guilt
  4. Ignore excuses
  5. Reinforcde sincerity, eye contact, First names
  6. If suspect cries, infer guilt
  7. Pose the 'alternative quewstion'- One more socially accpetabel still inferring guilt
  8. Admit guilt in front of witness
  9. Document admission 
9 of 14


Aim: Document a case of false confession of a youth who was at the time distressed and susceptible to interrogative pressure

Procedure: Case study of 17 year old youth accused of 2 murders of elederly women who had missing savings and been sexually assaulted; FC arrested because of inconsistencies in movement; No forensic evidence; FC confessed after 14 hours of agressive questioning and suggetsing he was sexually impotent;

Psychiatrists in prison evaluated him and found no mentail ilness, an dIQ of 94 and score of 10 on Gudjohnsson suggestibility scale

Conclusion: Case of 'Coerced compliant false confession'. He gave in to pressure to excape from the interview situation 

10 of 14

Creating a Profile

  • Top Down- Canter
  • Bottom Up- Canter
  • Case study of John Duffy- Canter
11 of 14

Canter- Top Down

Aim: test relaiability of organised/disorganised typologies 

Procedure: Analysis of 100 cases of serial killers from the USA; 3rd scrime commiyed by each serial killer analysed using The crime classification manual 

Results: 2x as many disorganised crime scenes were identified; 70% cases body was concealed; 75% sexual acitivyt occured; No further difference between organised and disorganised crime occured

Conclusion: Canter conluded that instead of their being a distcintion between two types of serial muder, all of the crimes had to have an organised element to them as they hadn't been caught after three killings. Better to look at personal variables 

12 of 14

Canter- Bottom Up

Aim: Identify a behaviour pattern from similiarities between offences

Methadology: Content analysis on 66 sexual offences from various police forces commited by 27 offenders; Subjected to Canter's smallest spave analysis

Results:  variables centeral to 66 cases:

  • Vaginal intercourse
  • No reaction to the victim
  • Impersonal language
  • Suprise attack
  • Victims clothing disturbed

Pattern= Attack is impersonal and suddena dn vicxtims response is irrelevant to offender

Conclusion: Five factor theory devised; Contribute to all sexual offences

13 of 14

Canter- John Duffy

Prelimanary Profile:

Residence: Lived in area circumscribed by first 3 cases since 1983; Lived in area at the time of arrest; Lives with a wife or girlfriend, no children

Age: Mid to late 20's; Light hair; About 5'9"; Right handed

Occupation: Semi skilled or skilled involving weekend work or casual labour from july 1984 onwards; Lack of contact with public character

Character: Keeps himself to himself; Very little contact with women; Knowledge of railway system along which attacks happened

Sexual activity: Variety of sexual actions = considerable sexual experience

Criminal records: Arrested some time between 24 october 1982 and January 1984 and this may have had nothing to do with **** but having been agressive under influence of drugs or alcohol.

John Duffy: Profile fit selection criteria; John duffy was interviewed by police for domestic violence against ex wife; Suspects connected via blood group; Duffy was a lot shorter than victims remebered and some describe dhim having black or ginger hair, Police suggest 'weapons effect' since he was carrying a knife

14 of 14


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Making a Case resources »