SPECIAL SEMINAR

 Wednesday, 9^th March, 2016

13.00 : Senior Common Room

Experimental Psychology, Priory Road

 Mr Alexander Etz

Visiting Scholar, University of Amsterdam

  “A Bayesian Perspective on the Reproducibility Project: Psychology”

 

Abstract

Using a Bayesian reanalysis of the recent Reproducibility Project: Psychology, we quantified the evidence provided by the two sets of studies (original vs. replication). After applying a correction for possible publication bias in the original results, we find that roughly 75% of studies gave qualitatively similar results in terms of the amount of evidence provided. However, most studies provide only weak evidence one way or another.

Biography

Alexander Etz's research includes a mix of development and application of Bayesian statistical methods, with secondary interests in both the history of statistics and advocacy for the adoption of Bayesian methods in the field of psychology. He completed his Bachelor's degree in psychology at The George Washington University in 2013, and he is currently working as a visiting scholar with Eric-Jan Wagenmakers at the University of Amsterdam.

His current projects include: How to quantify replication success, how to quantify statistical support for causal claims, how to account and correct for publication bias, and evaluation of how researchers interpret various levels of statistical evidence. In addition, he is working in tandem with the JASP statistical software team making a series of tutorial videos on how to perform and interpret Bayesian analyses using JASP. He also maintains a blog where he writes accessible tutorials on Bayesian statistical concepts.

 

ALL WELCOME