. The meta-analysis presented here derives the coding ex post facto through the
use of surveys of program administrators.
Instead of a meta-analysis interpreting
independent variables in relation to criterion main effect tests, the approach taken here was
to survey administrators using a standardized survey instrument.
This approach, merged
with the common criterion measure in a pre-post design, provides the potential for a meta-
analysis with less interpretive variation than that required with the use of published
research outcomes.