Measuring Impacts of Some Common Energy Conservation Programs Under Conditions of Nonrandom Assignment

Abstract

Nonrandom assignment to program and control group conditions can create serious difficulties for the evaluation of impacts of energy conservation programs. The particular analytic strategy that is appropriate depends on the character of the causal mechanisms associated with the nonrandom assignment. Guidlines for selection of the appropriate analytic techniques are presented and the analytical techniques described. It is recom mended that the possibility of performing a true experiment should be explored, beyond the use of random assignment to wait list and program conditions, in order to avoid the analytic complexities necessary in the nonrandom assignment designs.