although sample is large, the specific age group and culture means findings aren't representative of wider population so can't be applied to them = bad generalisability
overall strength as there is good internal validity and extraneous variables were controlled meaning study is valid despite the fact conclusions may be hard to come to
- peer ratings = labelling children as aggressive = could compromise children and raise protection from harm issues
- socially sensitive = concluded aggression is genetic = can create labels and self-fulfilling prophecy by identifying a 6 year old as genetically aggressive