Understanding the Uncertainty of Effectiveness-Cost Ratio in Educational Resource Allocation: A Bayesian Approach

• In Event: Bayesian Estimation and Propensity Score Analysis in Educational Statistics

Abstract

Despite the consensus on the importance of using cost-effectiveness analysis to guide decision making, whether it is sufficient to compare alternatives only based on one single, scalar efficiency ratio is still in debate. This paper synthesizes two perceptions of uncertainty, and evaluates several methods used to quantify the uncertainty of a ratio under each perception. Compared to other conventional methods, a Bayesian approach using a hierarchical model has at least two advantages. First, it provides a distribution of an efficiency measure to answer the questions that practitioners are most interested in. Second, the validity of the inference is not threatened by the limited number of observations available, a data feature that is commonly observed in educational cost-effectiveness analysis.

Author

• Yilin Pan, Columbia University