What JASP is
JASP is a free, open-source statistics package from the University of Amsterdam. It presents a point-and-click interface deliberately close to SPSS — load a spreadsheet, choose a test from a menu, tick options — with results appearing in tables already formatted to APA style, updating live as you change the options.
The feature that makes it more than an SPSS clone is that every standard test has a Bayesian counterpart in the same menu. A t-test and a Bayesian t-test sit side by side, take the same inputs, and produce output designed to be read together. That placement is a pedagogical argument as much as a technical one: it makes Bayesian analysis an option you can try on your own data this afternoon, rather than a separate discipline requiring a new toolchain. For a field arguing about p-values, that accessibility is the most useful thing anyone has built.
Why researchers use it
- It replaces SPSS for free — the same tests, a familiar interface, no licence, on Windows, macOS, and Linux.
- Bayes factors next to p-values — the most accessible on-ramp to Bayesian inference that exists, requiring no new software and no new syntax.
- APA-formatted output — tables you can read into a manuscript rather than reformat, which sounds trivial and saves genuine hours.
- Live updating — tick an assumption check or an effect size and the output changes immediately, which makes exploring options fast and their consequences visible.
- Better defaults than SPSS — effect sizes and assumption checks are prominent rather than buried, so students report results more completely with less prompting.
Where it fits in a research workflow
JASP handles analysis for studies using standard designs: group comparisons, correlations, regressions, ANOVA, and their Bayesian equivalents. Upstream is a cleaned data file; downstream is a results section.
It fits a specific and common situation — a researcher trained on SPSS, in a field where the standard tests cover the standard designs, who does not currently write code. For that person JASP is a strict improvement on the status quo: free, better output, and Bayesian methods newly available. Where it hands off is at the ceiling: the moment your model needs a custom contrast, a non-standard random effects structure, or a simulation, you need RStudio — and jamovi is the alternative that eases that transition by showing the R code behind each click. Preregister the analysis on OSF either way.
Getting started
An afternoon.
- Download it — free on all platforms — and open a CSV directly. There is no import wizard to fight.
- Run a test you already know, and read the output. The APA formatting and the prominence of effect sizes are the first thing to notice.
- Tick the Bayesian equivalent of that same test and read both together. The built-in help explains Bayes factors in place, which is the gentlest introduction available.
- The step people skip: save the
.jaspfile alongside your data and treat it as part of your materials. It records which analyses you ran with which options, which is the closest thing to an audit trail a menu-driven tool provides — and it is what makes your analysis checkable by a supervisor or a reviewer.
JASP vs the alternatives
| Alternative | Does it better | Pick it if |
|---|---|---|
| jamovi | Showing the R code behind every click | You intend to move to code eventually |
| RStudio | Everything, eventually — flexibility and reproducibility | The analysis is a contribution of the paper |
| SPSS | Institutional familiarity and legacy syntax files | Your department mandates it |
| GraphPad Prism | Wet-lab experimental designs and publication graphs | You are in experimental biology |
JASP and jamovi are close siblings. JASP leads on Bayesian methods and output presentation; jamovi leads on the path toward R.
Cost, licensing, and your data
Free and open source under AGPL, developed at the University of Amsterdam with academic funding rather than a commercial model. No licence, no account, no tiers.
Everything runs locally, so participant data stays on your machine — a meaningful advantage over cloud statistics tools for anyone working under a data agreement. The consideration is reproducibility rather than privacy: a .jasp file records your analyses and options, which is better than nothing and is not a script. It cannot be diffed, cannot be rerun from a command line, and cannot be composed into a pipeline. For a study where the analysis is routine, that is acceptable; for one where the analysis is the contribution, it is not.
The honest review
Strengths. JASP is the best on-ramp to Bayesian statistics that exists, and it achieves that simply by putting Bayesian tests where people already look. Beyond that, its output hygiene is a genuine improvement on SPSS: effect sizes, confidence intervals, and assumption checks are visible by default, and students trained on it report results more completely without being told to. That it is free removes the last reason for a department to keep paying for SPSS.
Limitations. It is menu-driven, and menus have a hard ceiling. Custom contrasts, unusual random-effects structures, simulation, and anything requiring a model the developers did not anticipate are simply unavailable, and the wall arrives without warning — typically mid-project. Data manipulation is weak, so cleaning happens elsewhere. Reproducibility is partial: the .jasp file documents choices but is not a script a reviewer can execute. And large datasets strain it.
Verdict. Adopt it if you are replacing SPSS, teaching statistics, or want to try Bayesian methods without changing toolchains. Skip it if the analysis is the intellectual contribution of your paper, where scripted RStudio is the honest answer. The condition that flips the answer is whether you will ever need a model that is not in the menu — and if you might, jamovi is the better bridge.
Common questions
Is JASP free?
Yes — entirely free and open source under the AGPL, developed at the University of Amsterdam. There are no tiers, no licences, and no account.
Can JASP do Bayesian statistics?
Yes, and it is the main reason to choose it. Every standard test has a Bayesian counterpart in the same menu, taking the same inputs, with built-in help explaining how to read Bayes factors. It is the most accessible route into Bayesian analysis available.
JASP or jamovi?
JASP if Bayesian methods and polished APA output matter most. jamovi if you intend to move to R eventually, since its syntax mode shows the R code behind every click and makes that transition gradual rather than abrupt.