Raine et al. (1997)

• To investigate if there were differences between violent offenders and normal people
• To see if people who had committed murder had brain dysfunction in: prefrontal cortex, amygdala, hippocampus, thalamus, corpus callosum, angular gyrus

• 41 people who had committed murder who were pleading NGRI
• 39 males and 2 females
• Average age was 34.3 years
• Had a range of different different diagnoses, which included organic brain damage (23), schizophrenia (6) and personality disorder (2)

• 41 participants matched on age and gender (independent groups design not matched)
• 39 males and 2 females 
• Average age was 31.7
• Screening process: their medical records were accessed, they were checked for general health, had a physical examination and a psychiatric interview (6 had shizophrenia). They were excluded if they had experienced if they had experienced head trauma, had a history of seizures or substance abuse

• Opportunity sample
• Were medication/drug free for two weeks prior to the study
• Gave consent before having the brain scan
• Carried out the same continuous performance task (CPT) which consisted of a sequence of blurred numbers which they had to recognise, while having a PET scan

• There was no difference in the performance on the tasks
• Lower activity (reduced glucose metabolism) in the: prefrontal cortex, parietal lobe (especially the left angular gyrus) and the corpus callosum
• Higher activity (higher glucose metabolism) in the: occipital lobe and the right side of the thalamus
• Asymmatrical activity (lower activity in the left and higher activity in the right) in the: amygdala and the medial temporal lobe including the hippocampus

• These brain differences have been associated with many behavioral changes that could be related to violent behaviour e.g dysfunction in the prefrontal cortex has been linked to impulsivity, lack of self control and an inability to learn from the consequences of behaviour
• The hippocampus, amygdala and thalamus have all been related to learning and it has been suggested that abnormal activity here could result in criminals being unable to modify their own behaviour by learning from consequences of their actions
• Findings from animal studies into aggression can be generalised to humans and there is a link between brain structure and aggression

• Independent groups design
• opportunity sample
• 41 participants
• Age and gender
• Medication free
• NGRI killers
• PET scan
• Glucose metabolism
• Lower activity - prefrontal cortex, corpus callosum

• Sample size was fairly large - 41 in experimental condition. More likely to be representative of NGRI killers
• Findings cannot be generalised to other types of violent offenders nor can they be used to explain aggression in the general population
• Conclusion: although abnormal brain activity may explain aggression in certain individuals, it is unlikely to be able to explain aggression for all people. More research is needed to look at aggression in the normal population

• Standardised procedure with objective measures of brain activity (PET scans) give good internal reliability
• Studies using PET scans have similarly found abnormal activity in these areas of the brain in violent, antisocial and psychopathic individuals
• Some variation in procedure - tricky to ensure that some parts of the brain are being scanned in different individuals. This reduces reliability
• Conclusion: scientific techniques and controlled procedures used in this type of laboratory study mean it would be regarded as having high levels of consistency and reliablity

Raine said brain scans could be used to inform decisions about the appropriate sentencing and to help determine how appropriate it would be to offer parole to prisoners 

• Use of comparison group, screening for things like epilepsy, using objective measurse of brain activity (PET scans) and controlling variables (e.g medication, CPT) increse the internal validity
• Extraneous variables such as social and situational factors (e.g the stress caused to the NGRI group by their trail for a serious offence) could have affected the findings
• Conclusion: the controls and objective nature of the research reduces the likelihood that the findings are affected by demand characteristics or experimenter bias

• Informed consent: since the sanity of the NGRIs is questionnable, it would not be possible to obtain informed consent from them. However, their lawyers had requested the testing and should have their best interests
• Socially sensitive because: 1. suggests it is in the nature of some people to kill, and therefore can't be held responsible. 2. individuals with such brains may be labelled and treated differently, despite never have committing a crime