I’m delighted to announce that PsychoPy has been awarded a Technology Development Grant from the Wellcome Trust for the purpose of creating online experiments. We had already released a proof-of-principle that this could be done, but this grant will give us the funding to finish the job to a high standard. Online experiments are going to become easy-to-create and easy-to-deploy and provide the full rich set of components you expect, all with the same dynamic real-time control that we provide in local studies.
On top of the fact that online experiments are coming, the fact that we’ll have a full-time programmer for the first time in PsychoPy’s history means that we should be able to take this project to a whole new level of professionalism. Hopefully you’ll see many fewer bugs and faster fixing of issues
We’re not releasing a schedule yet. There’s a lot of work to be done, but this grant will enable the following:
creation/completion of the PsychoJS library, porting the PsychoPy Python library into JavaScript
creation of a server that will operate as both
a host for launching online studies easily
an experiment repository for sharing your work with collaborators and/or public, with full (git) version control
adaptation of the PsychoPy Application to generate the JavaScript code and communicate with the server
better general support and bug-fixing for the current PsychoPy code
We’re very grateful to the Wellcome Trust for the money that is going to make it possible, and I struggle to state quite how excited I am about this new era of professional PsychoPy with online experiments. I hope you’ll share in that enthusiasm!
Make sure you turn on “watching” on the announcements section of https://discourse.psychopy.org so that you get updates (including job announcements when those are ready)
This is fantastic news. Congratulations! I humbly ask that you make it possible to load videos in online experiments that enable collecting reaction times.
Movies, sounds, reaction times will absolutely be in there. And the issues with movie/sound formats should go away (because browsers have the codecs built-in).
And with a full-time programmer we’ll be able to spend more time testing as well.
All this stuff will run on all devices (including iOS and Android). In time, we might even add support for video recording from your device’s camera!
This is truly fantastic news, congratulations to all PsychoPy contributors, especially Jon of course.
But I can find a grey lining in any silver cloud - I fear the prospect of needing to program in Javascript (where the Builder doesn’t do the job). I love Python!!
for local experiments you’ll still be able to use Python - we aren’t dropping support for that at all and, with a F/T programmer around the Python side will get better too
for web experiments, JavaScript is unavoidable but:
we’ll certainly try to auto-convert your Python code into JS and if your component is reasonably simple I think that should work fine
the JS code will look very, very similar to your current code snippets
PsychoJS will have the same classes with mostly the same attributes and methods as the PsychoPy lib. There are differences in the structural aspect of the scripts, which Builder will be handling for you, but at the local level the code is really similar.
Fantastic news! I think it will be really useful to have a one stop place for many different types of experiments (and written for scientists).
I’d love the option to collect free response data both online and for local experiments too (e.g. open-ended survey questions, essay text). We eventually got something working with wxpython but I think many people would find it useful
By the way, for a future online experiment my need would be to be able to launch automatically a video after displaying a first page with consent and instructions and having clicked on the “next” button. My concern is that the video must be read exactly the same way for all browsers, and online survey plateforms are not very good at doing that. Each video is followed by a question.
Is there a possibility to begin to do it by now with Psychopy or would you have some suggestions to do that alternatively ?
Many thanks.
Chris, you can’t do that now (we just got the grant, so we haven’t done the work yet!) but yes, we fully expect all that to be standard stuff for the new online platform.
May I please check if there is a course on Psychopy taking place anywhere, I am from London and starting my dissertation soon and I am very worried if I can learn Psychopy accurately enough to design my experiment, if there is a course I can join and learn Psychopy and if you can suggest then I would be very grateful.
