A similarity between the meta-analysis presented here and conventional meta-
analyses exists in the fundamental effort needed in coding program characteristics
(independent variables). Traditional surveys of published research outcomes need to
identify independent variables across studies as indicators of common underlying
hypothesis