The statistical evaluation design is a longitudinal cohort study involving large random sample surveys of young people (and their parents) in 10 EMA pilot areas and eleven control areas. Two cohorts of young people were selected from Child Benefit records. The first cohort of young people left compulsory schooling in the summer of 1999 and they, and their parents, were interviewed between October 1999 and April 2000 (Year 12 interview). A second interview was carried out with these young people between October 2000 and April 2001 (Year 13 interview). The second cohort left compulsory education the following summer of 2000 and young people, and their parents, were first interviewed between October 2000 and April 2001.
The report uses both propensity score matching (PSM) and descriptive techniques, each of which brings their own particular strengths to the analysis.