OS (e.g. Win11): PsychoPy version (e.g. 2026.1.1 Py 3.13): Standard Standalone Installation? (y) Do you want it to also run locally? (y) What are you trying to achieve?: The experiment is not terminating correctly, and I believe this is causing some data to not be saved correctly.
What did you try to make it work?: I have tried running it under different versions, enabling and/or disabling the escape key, adding a final routine that required a button press to force the end, adding a final routine that automatically timed out after one second. I’ve run the study locally to confirm it works, in principle, and also piloted it through Pavlovia with developer tools/event viewer enabled, and did not see anything that appears to explain this. Participants do not report any error message, just the end of the experiment.
What specifically went wrong when you tried that?:
Nearly every (~97%) of sessions on the Pavlovia dashboard shows as aborted, never finished, despite participants reporting they completed the full ~60–90 min study. Saved data is inconsistent: sometimes just a single row confirming the session started, sometimes more, occasionally a full/complete record. I have also gotten a different version of this experiment to work in a way that retained most of the data, despite the same high ‘aborted’ rate. I made copies of that experiment (while making sure to isolate them on GitLab’s end, so that they don’t share the same URL or file path) and simply edited the stimuli. All stimuli seem to be working fine, and the experiment runs locally in Pilot mode, and appears to terminate as intended. Nothing notable while monitoring locally in the browser’s developer tools log (have tried Firefox and Chrome). I’m recruiting participants through Prolific by publishing studies that direct them to the study’s URL: (ex. https://run.pavlovia.org/xxxxxxx). I’m using builder mode, as I’m not very familiar with python or JS.
Hi, thanks for replying. No code components within routines. The study presents audio (mp3) or audiovisual (mp4) stimuli in separate loops. All files are included as ‘additional files’ in the experiment settings’ online tab. Due to lag while presenting audiovisual stimuli, I added the ISI function to this loop, which should pre-load the video files during the fixation cross period. The file structure/contents appear to be the same both in GitLab and locally. I’m syncing using the browser sync button at the top of the builder, which does appear to update with the log files and csv sheets indicating, at least, that the participant started the experiment, and what occasionally appears complete data.
I’m currently investigating whether or not there’s something wrong with the stimuli itself that is causing a crash (although, the crash would happen at the end, during the ‘thanks for participating’ screen), and will update this post if I figure something out on this end.
I believe I’ve identified the issue as a problem with some of the video stimuli encoding needing to be changed, which caused an error (that only manifested at the conclusion of the study)? In any case, re-encoding these stimuli seems to have fixed the issue, so I’m not going to look a gift horse in the mouth. Is there a way to mark this thread as solved, or edit it to include what I did? Thanks for your help.