Reinforcing RCTs with Multiple Priors while Learning about External Validity

12/16/2021
by   Frederico Finan, et al.
0

This paper presents a framework for how to incorporate prior sources of information into the design of a sequential experiment. These sources can include previous experiments, expert opinions, or the experimenter's own introspection. We formalize this problem using a multi-prior Bayesian approach that maps each source to a Bayesian model. These models are aggregated according to their associated posterior probabilities. We evaluate a broad of policy rules according to three criteria: whether the experimenter learns the parameters of the payoff distributions, the probability that the experimenter chooses the wrong treatment when deciding to stop the experiment, and the average rewards. We show that our framework exhibits several nice finite sample properties, including robustness to any source that is not externally valid.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
11/02/2020

Bayesian sample size determination using commensurate priors to leverage pre-experimental data

This paper develops Bayesian sample size formulae for experiments compar...
research
01/30/2020

A Comparison of Prior Elicitation Aggregation using the Classical Method and SHELF

Subjective Bayesian prior distributions elicited from experts can be agg...
research
06/19/2019

From Local to Global: External Validity in a Fertility Natural Experiment

We study issues related to external validity for treatment effects using...
research
04/15/2021

Bayesian and Dempster-Shafer models for combining multiple sources of evidence in a fraud detection system

Combining evidence from different sources can be achieved with Bayesian ...
research
07/07/2022

A Pair of Novel Priors for Improving and Extending the Conditional MLE

A Bayesian estimator aiming at improving the conditional MLE is proposed...
research
07/29/2022

HUG model: an interaction point process for Bayesian detection of multiple sources in groundwaters from hydrochemical data

This paper presents a new interaction point process that integrates geol...
research
03/13/2022

Valid and efficient imprecise-probabilistic inference across a spectrum of partial prior information

Bayesian inference quantifies uncertainty directly and formally using cl...

Please sign up or login with your details

Forgot password? Click here to reset