In this paper I argue that research, development, and evaluation of social and educational programs should routinely be integrated through a planned variation approach. This approach typically involves the design, implementation, and evaluation of an enhanced, as well as standard, variation of a program. In contrast to conventional black box evaluations, I argue that a planned variation approach is better suited to the needs of program operators, offers methodological advantages, tests program theory, helps to build the base of knowledge for constructing better programs, and would foster greater utilization of evaluation findings. I discuss implications regarding the role and preparation of evaluators and issues in conducting planned variation evaluations. I compare the planned variation approach to other types of theory-based evaluation and discuss when and why it may be preferable.
Bibliographical noteFunding Information:
The author thanks Kathryn Boudett, Ana Cristina de Souza, Barbara Goodson, Timothy Hacsi, Tracy Huebner, Eliot Levine, George Madaus, Melvin Mark, Jack Needleman, Pamela Perry, Anthony Petrosino, Ana Yolanda Ramos-Zayas, Sean Reardon, and Carol Weiss for helpful comments on earlier drafts of the manuscript. This paper was made possible through support from the Spencer Foundation.
Copyright 2017 Elsevier B.V., All rights reserved.