| Session Title: Summative Confidence: The Mathematical Algorithm |
| Expert Lecture Session 351 to be held in Centennial Section B on Thursday, Nov 6, 3:35 PM to 4:20 PM |
| Sponsored by the Quantitative Methods: Theory and Design TIG |
| Chair(s): |
| Brooks Applegate, Western Michigan University, brooks.applegate@wmich.edu |
| Presenter(s): |
| Cristian Gugiu, Western Michigan University, crisgugiu@yahoo.com |
| Abstract: How can one estimate the precision of an evaluative conclusion? While confidence intervals (CI) are the standard method of examining the precision of a single variable, evaluative conclusions are formulated by synthesizing multiple indicators, measures, and data sources into a composite variable. Unfortunately, the standard method for calculating a CI is inappropriate in such cases. Moreover, no method exists for estimating the combined impact of sampling, measurement error, and design on the precision of an evaluative conclusion. Consequently, evaluators formulate recommendations and decision makers implement program and policy changes without full knowledge of the precision of an evaluative conclusion. This presentation will demonstrate how the Summative Confidence method can be used to estimate the impact of over a dozen factors on the precision of a conclusion. Therefore, evaluators will no longer need to assume that their conclusions are accurate. Instead, they can estimate the measurement error of each conclusion. |