Pumba is a chaos testing command line tool for Docker containers. Pumba disturbs your containers by crashing containerized application, emulating network failures and stress-testing container resources (cpu, memory, fs, io, and others).

JATOS greatly simplifies setting up a server to run studies online but you still need that server to begin with! If getting your own server (e.g. from your IT department) is not an option, you can set up JATOS on a cloud vendor (AWS, Azure, Google Cloud are the most common ones).

Matan Mazor suggested to use a cryptographic hash as a seed to randomize stimulus sequences in order to prove that the preregistration happened before data collection. I thought this is a good idea. You can easily do it with JATOS. Here is how.

There is a really easy way to embed Google Slides in your JATOS component’s HTML file. I thought that this might be a quick way to write the instructions for an experiment. They aren’t the most beautiful way to present instructions (the Google Slides frame is always visible) but it’s quick, easy, versatile, and allows your participants to move forwards and backwards on the series of instructions slides.

I set up a JATOS server where everyone can try out and test JATOS: https://cortex.jatos.org.

A JATOS user told me about Cloud9 and I thought I should mention it in a blog post.

Update: Unfortunatly Dply closed down and can’t be used any more to get an JATOS server.

Who reads descriptions? We mostly copy code examples and change them to our needs. That’s why I added many code examples in the jatos.js reference. I hope it helps!

Since JATOS doesn’t limit what actually runs in the browser during your studies you can use all kinds of JavaScript libraries and frameworks with it. This is not possible if you use e.g. Qualtrics. One of those libraries is p5.js. p5.js lets you do otherwise complicated 2D or 3D graphics in an easy way. Have a look at the examples on the p5.js webpage to get an idea of what is possible.

Doing user input forms, e.g. for questionaires or surveys in the browser yourself often means handling a lot of <input>, <textarea>, CSS etc. Those can be quite cumbersome. One wants to concentrate on the actual survey and does not want to count radiobuttons or the pixels of the border of input fields. Therefore usually one uses a library that helps with this process. So far there where mostly jQuery UI and Bootstrap. Alpaca also does the job but I found it a bit difficult to use (although I only tried for one afternoon). Now there is a new option: SurveyJS (or survey.js - both spellings seem to be used).

Since JATOS 3.1.6 it is possible to start JATOS as a systemd service. This is probably most useful for server installations. Everything one need to know is in the documentation.

We noticed that can be difficult for people who have never heard of JATOS to understand what JATOS actually is and what it can do for them. So we started with a little series of introduction videos. We planned 3 parts so far. First one is youtu.be/H1tR8gPt6bA. I hope you like it (I’m a programmer and not an entertainer - so please be gentle with me).

This post is about how much memory JATOS needs for a single user. If you intend to set up JATOS for several user (e.g. for an institute) JATOS needs more memory. The same is true for an experiment with many or large result data.

JATOS 3 is out and it got some nice security improvements that I want to talk about in more detail in this post.

It’s probably confusing for someone who starts with JATOS to grasp the differences between the JATOS worker types. So I made an overview table of them. It’s in the JATOS documentation: www.jatos.org/Worker-Types.html#overview.

Recently I was asked to help with a broken JATOS instance. The logs were full of error messages like

Update: Unfortunatly Dply closed down and can’t be used any more to get an JATOS server.