This package provides functions to test for a change of Bayesian evidence over time which can be used to examine volatile effects. It includes the following sequential Bayesian testing methods:
- Bayesian Sequential t-Tests
- Bayesian Sequential Non-Parametric Tests (Wilcoxon Signed-Rank Test & Mann-Whitney U Test)
- Bayesian Sequential Binomial Test
- Bayesian Sequential Correlation Test
Unusual temporal patterns of evidence accumulation can be tested against an empirical null distribution set via creating simulated data sets.
# Install remotes if necessary
if (!requireNamespace("remotes")) install.packages("remotes")
# Get package from Github
remotes::install_github("mrzdcmps/changeofevidence")
You can download 10,000 data files each containing 1,000,000 random bits generated by a quantum random number generator (Quantis by idquantique) as source for the monte carlo simulations here: https://osf.io/gs42z/files/