THE FALLIBILITY OF COMPARISON GROUPS

This article presents evidence strengthening the argument against comp arison group strategies given the current state of knowledge. The comp arison group for this study was drawn explicitly recognizing the self- selection issues. It drew youths from similar sources matched on attit udes related to California Conservation Corps (CCC) participation not usually measured and collected information explicitly designed to impr ove the statistical modeling of participation. Yet, as always, whether a comparison group strategy works is an empirical question. One belie ves the results from such a study only if they basically conform to on e's prior beliefs, otherwise the suitability of a comparison group is suspect.