Researchers can use sensitivity analyses to evaluate whether they, or stakeholders, should be careful in extending their findings on school choice programs to populations or locations beyond those studied.
An Exploration of Appropriate Sensitivity Analyses
ResearchPosted on rand.org Apr 1, 2016Published in: Journal of School Choice: International Research and Reform, v. 10, no. 1, Mar. 2016, p. 48-72
School choice programs continue to be controversial, spurring a number of researchers into evaluating them. When possible, researchers evaluate the effect of attending a school of choice using randomized designs to eliminate possible selection bias. Randomized designs are often thought of as the gold standard for research, but many circumstances can limit external validity of inferences from these designs in the context of school choice programs. In this article, we examine whether these limitations are applicable to previous evaluations of voucher, charter schools, magnet, and open-enrollment programs. We devise simple sensitivity analyses that researchers could conduct when analyzing lotteried programs to determine whether there are reasons to be cautious about the breadth of appropriate inferences.
Researchers should evaluate whether their findings are externally valid by using the sensitivity analysis tools discussed in the article.
This publication is part of the RAND external publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.
RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.
