| Session Title: Prescriptive forecasting: Putting Our Programs and Policies Into the Proper Context |
| Demonstration Session 501 to be held in Panzacola Section H2 on Friday, Nov 13, 10:55 AM to 11:40 AM |
| Sponsored by the Quantitative Methods: Theory and Design TIG |
| Presenter(s): |
| Patrick McKnight, George Mason University, pmcknigh@gmu.edu |
| Abstract: Prescriptive forecasting allows evaluators to use cost-effectiveness and probability benchmarks to guide program development and evaluation. Program and policy developers frequently begin with little guidance concerning costs, outcomes, or efficiency. If developers and evaluators used informative benchmarks that were quantifiable then our outcomes might be more easily conveyed. Cost-effectiveness studies, Bayesian probability models and effect size estimates are essential methods for prescriptive forecasting. Cost-effectiveness tools help shape both the cost of a program along with the expected outcomes. Bayesian models help refine those parameters so evaluators may see how different program aspects (e.g., costs per unit served, time required for full implementation, and stakeholder support) may affect these relevant outcomes. I plan to present an example of prescriptive forecasting using these quantitative tools. My aim is to show that the process of setting developmental benchmarks requires little skill and can result in easy-to-communicate results. |