JASP: promising free statistics software for Macs

JASP was created as “a low fat alternative to SPSS, a delicious alternative to R,” and comes out of the University of Amsterdam. It is, in essence, a graphical user interface for the free, but hard to use, R statistics program.

Updated Sept. 2024 by David Zatz

JASP is a Linux native, but runs well on Macs using XQuartz windows. It uses the native Mac open/save dialogue box, once you click on the menu icon, then Open, then Computer, then Browse; and it’s easy to install, unlike, say, PSPP.

The software looks and feels something like SPSS, and feels more native than SPSS itself (which is native). Calculations and screen drawing are far, far, far faster than in “real SPSS” — when you select the tests, they might actually be pumped out before your finger is fully off the mouse. When you make changes, they occur instantly—so if you need a record of what you’re doing, be careful!

We loaded our big test file quickly, and ran descriptives instantly. When it could not finish one process in a reasonable time, we ran others, indicating it’s well multi-threaded. The survey file also loaded in a fraction of a second; frequencies for all 40 five-point variables came in the blink of an eye. Opening very large SPSS-format files can take a little more time, but if you save them in JASP’s own format, the open nearly instantly.

JASP

There are some strange aspects of JASP for users of other software. First, the program has a “three column” setup; it starts out just showing the data view. When you do an analysis, first it shows you the options; then, once you start selecting variables and such, the actual analysis is done in real time on the right side of the screen. If you make changes, they are instantly applied to the analysis. Once you start doing analyses, though, to get back to the data, you have to drag a vertical bar on the left side of the screen over. Then you can show all three columns.

For t-tests, when Levene’s test is significant, SPSS and PSPP use the "unequal variances" formula for Student's test; JASP and Jamovi want you to switch to Welch’s test, which makes fewer assumptions. They don’t actually tell you that, they just have a footnote saying you’ve violated the equal variance assumption, though Student’s t-test actually does have a formula for groups with different variances. (Now the warning is quite subtle in JASP and doesn’t suggest any alternatives.)

Survey researchers will be happy to know they can assign value labels — and unhappy to know they must be done variable by variable, without syntax. The labels used to be retroactively applied to whatever is in the output window, very rapidly, but now don’t seem to appear there at all.

There is a copy function for either individual charts or whole analyses. Copying worked just fine for graphics, and there’s also an easy graphics export.

jasp t-tests

Recoding is rather clunky, but it does work, and has improved since past versions.

JASP now lets you assign missing values to individual variables, though it can be weird—it wanted us to use the labels, not the values, in one case.

Independent samples t-tests require a variable with two values only (you can use missing values to acheive this). You can’t simply pick the values you want as you can in SPSS.

Thanks to Prof. Kim-Oliver Tietze for pointing us to JASP.

JASP project/download page • Other free Mac statistics packages • PAST • Jamovi