Abstract：Over the past decade psychologists became to realize potential problems with the statistics using conventional null hypothesis significance testing (NHST). The NHST has been misused in many questionable research practices. To overcome this issue, recently researchers propose to adopt Bayesian statistics, such as Bayes factor methods, instead of the NHST approach. A Bayes factor is a ratio of the likelihood probability of two competing models (or hypotheses), usually a null and an alternative. Here we highlight a number of advantages using Bayes factor over conventional NHST, and illustrate how to calculate Bayes factor using the state-of-the-art R packages of Bayesfactor and the software JASP. Finally, we listed some potential problems in the application of Bayes factor that we should be aware of.
浙公网安备 33010602002302号 备案号：