Mark Merolli
More and more often, randomised controlled trials are used to assess how well interventions work to improve health. Individuals in clusters are frequently more similar than those in separate clusters, regardless of therapy, which is a crucial characteristic of CRCTs. When designing CRCTs to achieve acceptable sample sizes and when analysing clustered data to obtain accurate estimates, it is important to take into account this similarity within clusters. Methods: Year olds' nationally representative data from the nations that took part in the 2007 European School Survey Project on Alcohol and Other Drugs were examined. For substance use and psychosocial health, intra-class correlation coefficients at the national school level were determined.
English 
Spanish 
Chinese 
Russian 
German 
French 
Japanese 
Portuguese 
Hindi