Supplement for “ Bayesian Hypothesis Testing for Single - Subject Designs ”

This document provides example R code demonstrating how to use the BayesSingleSub package (Section 1), and the technical details for the sampling routines (Section 2). 1 Tutorial for computing de Vries and Morey's Bayes factors Here, we show how to compute the Bayes factors B ar , B trend , B int , and B t+i , and how to obtain and plot the posterior distributions of the model parameters. First, download the R statistical environment from http://cran.r-project.org/ and install the BayesSingleSub package using the R command: install.packages("BayesSingleSub") Then, load the BayesSingleSub package with the library() function: library(BayesSingleSub) For the purposes of this demonstration, we compute the Bayes factors for the data shown in Figure 1 of the manuscript. We first define the data and the number of observations in the pre-and post-treatment phases: For convenience, we divide the data before and after the intervention into separate vectors: ypre = data[1:n1] ypost = data[n1 + 1:n2] The logarithm of the JZS+AR Bayes factor B ar can be obtained by using the ttest.Gibbs.AR() function, and the logarithm of the TAR Bayes factors B int , B trend , and B i+t by using the trendtest.Gibbs.AR() function: