Evaluate

• Culturally biased

  – only used children from Roman Catholic schools, ignoring other religions or children without religious beliefs. Religion has a strong influence on children’s moral behaviours.

  -As the school was independent, all students had to pay a fee, meaning most would come from better off families. Income and crime are often linked so it is a limitation that poorer children were missing from this sample.

  -represent patterns of behaviour in Australia; we cannot assume that other cultures would have the same results.
• Age bias 


  Average age of 14 and then 16 (time 2), it may be they were too old to properly establish the factors that lead to delinquency. This is because behaviours such as interpersonal violence, vandalism and theft almost definitely start much earlier than 14-16 for children who are going to get involved in delinquency.
• The 20% of participants who dropped out of the study by time 2 may have biased the results 


  Participants who dropped out may have represented particular types of people more than others. It could include people who engaged in lots of criminal activity by 16 and did not want to report it or people whose self-esteem was so low they did not want to continue with the study. This would affect the overall validity of the study.
• The use of self-report can lead to invalid data

  due to social desirability bias, the participants may not have been honest about information such as how much delinquent behaviour they have been involved in or how low their self-esteem was. Self-report rely on people’s insight, it might be the participants were not that aware o their personality when answering questions.
• The use of closed questions can be criticised for their lack of construct validity 


  Using simple options (rating scales) to measure complex constructs (personality and delinquency) is not appropriate. Using quantities data hides the depth of the relationship between different factors.