Data collection in Pavlovia

Hey everyone,
I piloted today my online experiment created in jsPsych, hosted by Pavlovia and acquired participants through Prolific. I have the problem, that when I click on “Download Results” I only get a fraction of the results that should have been submitted, but to me it definetly looks like the participants completed the experiment correctly since they got redirected at the end of the study.

Anyone who had the same problem or knows what I could do about this?

Many thanks for your help and best wishes!
Linus

Hey Linus,

Not sure what could be the problem, but I can say that Pavlovia checks quite carefully whether all submitted data has been received and stored before sending participants onwards. Maybe some issue in jsPsych?

Best, Thomas

Honestly I have no idea what could cause the problem and it makes me even more confused that it worked in a coupe of cases but in most not… when i test the experiment on running data collection works just fine. You have any idea how I could test for the problem?
At least one participants contacted me and said that at the end of the experiment it said “submission failed” but I don’t know where that should come frome

Best Linus

Ow that’s a tough one! Here are some reflections and pointers that might help:

  • It could be that your experiment doesn’t work on specific operating systems and web browsers. That’s hard to examine for yourself. I do know the jsPsych team does some testing on this, but perhaps something special happens in your experiment that they don’t include in their tests.
  • Could help to reach out to the jsPsych team, feel free to put me in the loop, we’ve got a good relationship (on GitHub I’m tpronk)
  • You could consider extending your experiment such that besides submitting the data to Pavlovia, it offers the data for download, and ask your participants to e-mail the data to you.
  • Finally, the more detailed bug-reports you can get from your participants, the better. I wrote a little guide for that: https://thomaspronk.com/pdf/Tech_Support_and_Bug_Report_Guidelines%20v1.3.pdf

Thanks for your reply!
So I did not yet find out what the problem is but there is definetly one in the interaction of Pavlovia and jsPsych.
I’ve put up a super simple “experiment” with just 3 questions and sent it to my friends to complete it and even there are issues: Pavlovia is receiving far too many requests for my experiment, consuming the credits!
This is super weird (see screenshot the last 6 entries) it received 6 participations of the experiment within some seconds distance but that was none of my friends (and no, it was not just one person opening the tab 6 times, I asked them).

Screenshot 2021-03-31 151311|690x197

I would open a thread on github but I have never done that before so I am not sure how and where to put it to adress the right people, I would be happy if oyu could help me with that :slight_smile:

Thank you or anyone very much!

Sounds like a concurrency issue. JS is structured as a set of functions calling each other, but if a function is called too often then this kind of stuff can happen. Maybe a good first step is to add some debugging code to see if (and when) that happens. See this tutorial: Thomas Pronk / tutorial_js_console_log · GitLab

No worries about the jsPsych group; it’s not like a formal bug report or a Pull Request, more like a forum such as this one. Discussions · jspsych/jsPsych · GitHub