JASP 0.8.6 Preview: Frequentist Chi-Square Test and Revamped Bayesian Regression

Soon we will release JASP 0.8.6. In this post we preview three important new features.

1. Frequentist Multinomial and Chi-Square Tests

JASP 0.8.6 provides frequentist multinomial and chi-square tests. For count data, users can test whether the cell frequencies are equal (i.e., the multinomial test), but users are free to specify any pattern of expected frequencies (or probabilities) and test this against observed frequencies. The expected frequencies can be entered manually or through the drag-and-drop interface. Below is a screenshot of the chi-square test in action:

2.Revamped and Greatly Expanded Bayesian Linear Regression

JASP 0.8.6 presents a revamped Bayesian Linear Regression routine. Previous versions of JASP used the BayesFactor package and focused purely on Bayes factors. The upcoming version has implemented the functionality of the BAS package from Merlise Clyde (we are grateful to Ioannis Ntzoufras for suggesting that JASP includes the BAS functionality, and to Merlise Clyde for making some BAS adjustments that made our job easier). As you can see from the vignettes here, the BAS package allows users to explore posterior distributions, examine residuals, and conduct multi-model inference. In addition, the BAS package presents a veritable cornucopia of different model specifications. Finally. BAS is remarkably fast. Of course we still offer the functionality of the BayesFactor package — in fact, it is the default option. Here’s a screenshot showing some of what the new BAS-style regression has to offer:

3. Easy Install on Most Linux Distributions

JASP 0.8.6 supports Flatpak so that it can be easily installed on most Linux distributions. Previously only Ubuntu was supported out of the box, but from now on users of Fedora, Debian, Arch, or other Flatpak-supporting distributions can install JASP without hassle. We will also try to add JASP to the public flathub.org repository to make it as simple as possible to run JASP on your favorite distribution.

JASP 0.8.6 presents many other improvements (e.g., the frequentist Z-test, simple main effects for repeated-measures ANOVA, greatly improved interface for high-resolution screens, etc.) and we will provide a detailed list in next week’s post.

We are excited about the upcoming 0.8.6 version and we hope you are going to like the new functionality.

Like this post?

Subscribe to our newsletter to receive regular updates about JASP including our latest blog posts, JASP articles, example analyses, new features, interviews with team members, and more! You can unsubscribe at any time.

About the author


We're the JASP Team!