Possible bug: experiments ran online on Pavlovia show only one trial then skip to next routine

Online experiments
,
1

Good morning!

I would like to report a possible bug that I incurred in on psychopy version 2024.2.2win64, standalone with python 3.8.

I created an experiment in builder, a simple go/nogo inhibition task. It works perfectly in the builder, but I plan to use it on Pavlovia for an online study.

I succesfully transfered it on my Pavlovia Dashboard. Everything runs well up until past the “instructions” Routine.

Essentially my flow is composed of two parts: a practice Routine with feedback, and the experimental Routine with the actual experimental trials with a “block” loop. Each Routine has the same “conditions.xlsx” file, with 8 conditions (6 go trials and 2 nogo trials, that I repeat multiple times so that I have multiple of 8 trials but that’s beside the point).

The instructions page runs smoothly, however from that point on it shows only 1 condition per part and then moves on to the next (i.e., 1 trial for the practice, then skips immediately to the instructions for the experimental trials. Past that, it shows 1 experimental trial, then since I repeat the block once it skips immediately to a “pause” routine. Past that, it repeats by showing only 1 experimental trial and then immediately skips to the “goodbye” Routine message).

This behaviour happens regardless of the experiment being marked as “piloting” or “running”.

I scratched my head a bit thinking what I did wrong, so I decided to download a demo gonogo task from the “Explore” tab in Pavlovia. I cannot post the link because I’m a new user and I am limited to two links (used down below). To access this Demo, go to the Pavlovia website, then click on Explore, then scroll all the way down to Public Demos and you should see a demo called “go_nogo” with a green “tick” and a red “X”, made by Rebecca Hirst.

I uploaded it on my experiments on Pavlovia and I ran in the same exact issue. It’s only one block, so it shows only 1 trial, then skips to the goodbye message, where it says to press “space” to leave the experiment, although nothing happens and I remain stuck there (I have to press “esc” and close the tab abruptly).

I thought that maybe it was a Psychopy issue, so I downgraded my version to 2024.1.5win64 (don’t know the python version because there was only that one option on GitHub).

I reuploaded both my experiment and the demo experiment under that version and they both ran smoothly.

Here are the link for the gonogo demo under version 2024.2.2 and under version 2024.1.5. The repositories should both be public. Needless to say, I didn’t apply any change to the code except for the Psychopy version it was “built” in.

I would have tested it under other Psychopy versions, but my laptop runs on coal and steam and it would take too long.

I hope I sufficiently explained the problem and I remain available for any further question.

Thank you in advance!

2

Routines skipping is usually due to the response that ends the routine (e.g. a mouse press) ending multiple routines at ones.

However, in your case it’s due to a bug in 2024.2 with non-slip timing (green routines of a known duration) which I thought had been fixed. There is a workaround of removing the durations in the routine and instead adding a duration in the routine settings.

1 Like
3

Thank you!

I think for now I will be working with 2024.1.5, because I am building a large number of “basic” tasks for my lab so that they can copy and adapt them for their own particular variations.

I could include the workaround, but I’d rather keep things simple and not have to come back and modify the code for a lot of experiments (it’s probably not a difficult workaround, but in case I realise something is wrong in the way I built tasks I’d rather reduce the time I spend going through each and apply the corrections).

In any case, it’s good to know if any other user incurs in the same issue and needs a solution.

Cheers!

1 Like