Studies in detail Criminal Psychology

Aim: To investigate the effect of introduction of satellite T.V on the aggressive behaviour of children

Background: aggressiveness measured in 1994 prior to the introduction of T.V in 1995

Method and Procedure:

• naturalistic experiment on 3-8 year olds on St Helena island who had not seen T.V
• children from two schools (160 pupils in each school) were filmed in free play in the playground
• 26 playground behaviours operationalised: kicking, pushing and hitting. Prosocial behaviours too: hugging, affection, sharing
• a pair of independent coders scored each 60 secs of video

Results

• low level of antisocial behaviour
• the mean number of each behaviour was compared in cross-sectional analysis between data from 1994 and 2000
• no significant differences in antisocial or prosocial behaviour
• initial level of aggression was very low and remained so as viewing opportunities increased
• children displayed twice as much prosocial behaviour
• boys likely to be anti social and girls more likely to be prosocial
• 

Conclusions

• exposure to more violent T.V does not always result in more aggressive behaviour. Therefore TV has no influence on behaviour
• 
• Culture affects behaviour- tight community so parents strict with children.

Evaluation

• made sure thet filmed after film crew had stopped attracting attention- less chance of demand characteristics
• recording not conducted in bad weather- less chance of extraneous variables
• independent coders watched the video of the children- less chance of experimentor biased
• Large sample
• naturalistic observation- less chance of demand characteristics, ethical situation
• Camera visible to children so high chance of demand characteristics
• Not generalisable- no urban culture, more of a case study so unique situation
• Social factors (community) not taken into account.

Aims:

• To see if leading questions are likely to affect someone's response
• to see if the phrasing of a question affects estimates of speed

Method/Procedure:

•  45 student participants put into groups and each shown films, each involving a car accident
• some participants asked: 'How fast were the cars going when they hit each other'
• others: verb replaced with: bumped, collided, smashed, contacted.

Results:

• Smashed gave the highest estimate of speed and contacted was the lowest
• 

Conclusions:

• Wording of questions does have an effect on eyewitness memory.
• 
• Verbal labels, such as ‘hit’ and ‘smashed’ had a major effect on eyewitness memory, and it was suggested that schemas came into force, modifying the memory to better fit the verbal labels supplied after the incidents.

Evaluation:

• Labatory experiment with clear controls- less chance of confounding variables affecting results .
• Study replicable so therefore reliable
• lots of quantatative data meaning research is valid and can be analysed objectivly- less chance of experimentor biased
• helps police officers in questioning techinques and led to interviews being recorded

• results led to cognitive interview less people wrongly convicted as police are improving witnesses memory

• student participants not under strain of a real accident so lack of ecological validiy
• participants pay more attention to a video than they would a real accident as they had fewer distractions- less validity
• co witnesses have an influence at a real incident
• participants all students so not generalisable to the whole population- students used to remembering things so have an advantage. and age may have meant they have better memories
• Lab setting- more chance of demand characteristics
• make of car may have has an effect- sports car you would estimate going fast
• students may not have driving experience so found it harder to estimate speed.

Aim and Background:

• Vancouver, Canada a theif stole money from a gun shop then shot at the owner.
• The owner then fired and killed the theif
• witnesses viewed the scene from different locations

Yuille and Cutshal chose this case because there were alot of witnesses to compare results.

Aim: look at the problems of laboratory research in studying eyewitness testimony, to look at the accuracy of eyewitness accounts

• to compare eyewitness accounts taken straight after an incident with those taken four to five months after.
• to see how eyewitness memory could be affected by leading questions.

Method and Procedure:

• 21 witnesses to the crime
• police interviewed witnesses immediately after and wrote down exactly what they said then asked questions
• 4-5 months later 13 of the 21 were interviewed by researchers- they were audio taped and transcribed
• researchers got their own account then asked them questions just like before
• added 2 leading questions: 1) Did you see a busted headlight? or 'Did you see the busted headlight?' (there was NO broken headlight)
• 2) Did you see the yellow quarter panel? or Did you see the yellow quarter panel? (quarter panel was blue)
• researchers asked about degree of stress each witness experienced at the time of the accident
• 7 point likehurt scale 1 being calm-7 being anxious

Results:

• of the 13 witnesses, 7 central to the action and 6 peripheral. both were equally accurate
• in police interviews, those central 84.56% of central witnesses accurate compared to 73.32 of peripheral group
• even after 4-5 months details were still high and aaccurate
• leading questions had little effect (10 said no headlight or no yellow panel)
• 

Conclusions:

• EW are not inaccurate in their accounts
• Y and C suggested may have experienced falshbulb memories- those directly involved remembered more
• Shows just because some minor details may be wrong, it does not mean central details are wrong

Evaluation:

• they were well informed as 8 dropped out and therefore were able to withdraw - high on ethics
• field study- real incident with real witnesses so high E.V
• real life application- shouldnt dismiss an EW on minor details (EWT can be trusted), helps with testimoney if police use verbation interview
• witnesses treated excatly the same as the police treated them, same interview technique. High ecological validity
• random ages so generalisable
• little sample so not representational (but cannot be helped)
• quite a high number of witnesses to the crime but cant generalise to target population as people view incidents differentl
• one off case so not reliable or generalisable
• did not take into account witnesses may have remembered more due to high level of adreneline rather than reconstructive memory

Aim:

• to look at the effect of disguish on misidentification
• to see if instructions to review an incident (given before the incident) would affect identification in a line up
• to see if the time gap between identification, would affect it

Procedure

• 215 male, 375 female aged from 18-70 in a shopping mall were approached by a woman (the target) who asked either for directions and help finding jewellery
• 2 mins later, participants approached by a female researcher and asked to take part in a study on perception and memory
• given 16 question recall test that included questions about physical characteristics and clothing of the female target
• then given 6 photographs and asked to identify her. Informed that female might not be in the set
• debriefed at end of the study

Results:

• 49% of participants correctly identified her in the line up
• when woman was not in line up, 62% correctly said she was not there
• those that were told by the target that there would be a memory test did better on the questionnaire test but not in the line up
• 

Conclusions:

• about 50% of the time correct identification made
• witness preperation did not improve EWT
• jurors are likely to overestimate an EW ability to recall and to pick a person from a line-up so EWT needs to be reviewed.

Evaluation

• supports Haber and Haber (2001) which increases reliability
• field experiment- high E.V because participants in a realistic environment
• high E.V as witnesses not aware of the study until approached so would not have been paying particular attention to the target
• study has good generalisability as sample is typical of real witnesses due to random ages and backgrounds
• should we rely on EWT in court? helps with the review of this
• some of the female targets wore disguise which increases E.V
• Findings are reliable-  precise scoring procedure. If this procedure were to be repeated, then the study is likely to produce the same results.
• The data produced were quantitative, and so researcher bias is limited and is unlikely to affect the findings.

Evaluation cont:

• real witnesses asked about an incident which is more emotive than remembering a woman asking you something
• no controls on situational varibles- recall may have been affected as participant was too busy so paid less attention.
• in a real incident, the crime is quicker so reall witnesses less chance to take details in
• may not be generalisable- unfair to criticise lab experiments with the evidence from this study. as the study assessed real witnesses from a very small sample of people.