Twenty-one studies met the inclusion criteria (18 RCTs and 3 overviews).
1. Effect sizes in 6 RCTs ranged from 0.3 to 1.3, indicating that all the parent-training programmes reviewed were effective in producing positive improvements in parental perceptions of their children's behaviour.
2. Group-based parent-training programmes have a positive impact on the behaviour of children between the ages of 3 and 10 years.
3. Group-based parent-training programmes are more successful in improving the behaviour of children compared with methods that involve working with parents on an individual basis. This finding was consistent across both parent-report outcome measures and, to a lesser extent, independent observations of children's behaviour.
4. Parent-report outcome measures also showed that community-based group parent-training programmes produced more changes in children's behaviour than individual clinic-based programmes, and that community-based programmes may be up to six times as cost-effective and more acceptable to many parents.
5. There is still insufficient research to demonstrate which aspect of group parent-training programmes is the decisive factor in bringing about change. The only study using a placebo control group in which parents presented and discussed their concerns about parenting with other group members, and in which no set parent- training curriculum was used, showed that there were no significant differences in the results between the placebo and the treatment group.
6. While all group-based programmes produced changes in children's behaviour, the more behavioural type of programme in which the parent was trained to use reinforcement effectively, appeared to produce the best results when compared with Parent-Effectiveness Training (PET) and Adlerian programmes.