An evaluation method in which children referred to a program or other intervention are compared with a group of matched children who do not receive the program. For example, if a program is offered to children in one geographical area, their progress might be compared to that of similar age children living in another geographical area where the program is not offered.
The prefix 'quasi' indicates that such evaluations have some but not all of the attributes of a fully experimental or randomized controlled trial. There is a comparison group, but it is not selected at random. As a result, quasi-experimental evaluations are prone to selection effects. For example, the results of a quasi-experimental evaluation that compares children in one geographical area (receiving the program) with those in another (not receiving the program) may reflect the difference in demographics and not a relative measure of effectiveness.
Quasi-experimental evaluations have been shown to produce more optimistic results than randomised controlled trials of the same program. Specialised statistical techniques can reduce selection effects.