What Works for Whom? A Bayesian Approach to Channeling Big Data Streams for Public Program Evaluation
Publisher: American Journal of Evaluation (online ahead of print, subscription required)
Nov 19, 2017
- The novel Bayesian approach presented in this article produces better inference than the standard approach, and does so both more efficiently and without sacrificing crucial methodological principles.
- Bayesian adaptive design benefits study subjects by assigning them to more effective treatment arms.
- Bayesian adaptive design benefits evaluators performing the study by allowing for smaller and more informative studies.
- The benefit to study subjects may be attractive to practitioners who often have concerns about the ethical implications of denying services or staying with a treatment that is not effective.
You may also like...
A New Role Emerges for Principal Supervisors: Evidence from Six Districts in the Principal Supervisor Initiative
States and Districts Can Make Greater Use of Kindergarten Assessments to Understand Students’ Knowledge, Skills, and Growth