Bayesian data analyses

UBCO

A First Steps Guide To The Transition From Null Hypothesis Significance Testing To More Accurate And Informative Bayesian Analyses

Reference:

O'Connor, B. P. (2017). A first steps guide to the transition from null hypothesis significance testing to more accurate and informative Bayesian analyses. Canadian Journal of Behavioral Science, 49(3), 166-182. https://dx.doi.org/10.1037/cbs0000077

Abstract:

This article begins with a brief summary of the problems with null hypothesis significance testing (NHST), followed by a short, nontechnical description of perhaps the most useful NHST alternative, Bayesian methods. Simple R commands and output for Bayesian correlations, regressions, and ANOVA are provided. This is followed by examples of how to describe Bayesian analyses in the Methods and Results sections of articles. The focus is on taking the cautious first steps in a transition away from NHST.

Click here for the R code and output for the Bayesian data analyses that were described in the above article:

The links below provide R code and output for Bayesian analyses of the example datasets that are used in the following textbook:

Field, A., Miles, J., & Field, Z. (2012). Discovering statistics using R. Los Angeles, CA: Sage.

The R code below was produced in collaboration with Dylan Ermacora.

Chapter	R code	R output
Chapter 6: Correlation	Ch6_Correlation.R	Ch6_Correlation.txt
Chapter 7: Regression	Ch7_Regression.R	Ch7_Regression.txt
Chapter 8: Logistic Regression	Ch8_Logistic.R	Ch8_Logistic.txt
Chapter 9: Comparing two means	Ch9_Two_Means.R	Ch9_Two_Means.txt
Chapter 10: Comparing several means (ANOVA)	Ch10_ANOVA.R	Ch10_ANOVA.txt
Chapter 11: Analysis of covariance	Ch11_ANCOVA.R	Ch11_ANCOVA.txt
Chapter 12: Factorial ANOVA	Ch12_Factorial_ANOVA.R	Ch12_Factorial_ANOVA.txt

Brian P. O'Connor
Department of Psychology
University of British Columbia - Okanagan
Kelowna, British Columbia, Canada
brian.oconnor@ubc.ca