Description of the problem:
We ran 5 subjects through Prolific in the above task just to pilot and make sure everything is working. Everything seemed to go well, except that when I updated the git repository on my computer, there were 7 data files, and 1 participant seemed to be missing. When I check the Pavlovia dashboard it says 5 completed and 3 aborted, and when I download the data from the dashboard, indeed the zip file contains 8 data files. Why then are there only 7 data files in the gitlab repository for the experiment?
I found a similar post from a year ago, Data not saving through Gitlab but there are no responses on possible causes for the discrepancy.
Description of the problem: New data is not being reflected on the gilab page for one of my experiments. I thought that data was not being saved at all; however, if I download directly from the pavlovia website (using the download results bottom) I do see it.
I don’t think so. How can I check? It was saving the files correctly on the gitlab until yesterday. The only change that I did to the task was to remove some files from the local data directory.
Hi! I am having this issue too. I have a two part experiment (two separate sections) the first section saves a folder called “data” in gitlab, the 2nd does not and I can only download that data from pavlovia (so the participant files are being saved but for some reason not being sent over to gitlab for this part of the experiment). I am not quite sure why this is happening. Anyone find out why?
I have one experiment called ASLTestBattery-1 which I am able to download data from both pavlovia and gitlab. I have another experiment called ASLTestBattery-2 in which I am only able to get the data files from pavlovia. The gitlab does not have a “data” folder for this experiment.
Yes they are separate urls. ASLTB1 is totally fine and saves what it gathers for that experiment to gitlab. ASLTB2 does not save the data it gathers to gitlab. how do I get asltb2 to do what asltb1 is doing? the Results format is set at CSV for it just like asltb1 is
It sounds like there might be an issue with the syncing between Gitlab and the Pavlovia runtime server. Please could you let me know the full URL for asltb2 and I’ll ask Alain if he can take a look.
I am running an online experiment, but I’ve noticed a discrepancy between the data available in my Gitlab repository and the results when I manually “download results”. It seems there is an issue with data synchronization between Pavlovia and Gitlab, leading to missing data files in git (especially audio files). I am working on automating a regular cloning of data from Gitlab to my local system, but this disconnect is causing challenges.
Has anyone experienced this before and found solutions?
To clarify – do you have a microphone component?
Audio files from a microphone are not saved in the Gitlab repository and are therefore only available when you manually download results. Audio files from local runs will appear in the data folder and be pushed online when you sync.
Hi, I was just noticing a similar problem (although not with audio files) and figured I’d reply here.
I’ve discovered something quite strange. Initially, it looked like ~30–40% of participant data were not saving in gitlab and I was really concerned. Specifically:
Many CSV files were missing from the GitLab repository (where I typically sync/download data).
For some participants, there was no CSV file at all.
For others, CSV/log files appeared empty.
Pavlovia was also reporting ~40% of sessions as “aborted.”
However, I’ve now realized the if I use the “Download Results” button directly on Pavlovia, the previously “missing” data are actually there, and the files appear intact and complete.
Additional context:
All participants are being redirected to Prolific with the correct completion code.
There is no custom JavaScript (I used only standard Builder code components).
I’m using PsychoPy version 2025.2.4.
The task is time-locked (~60 minutes minimum), so very short completions are not possible.
The correct number of credits were used.
I wonder why data would be accessible via the Pavlovia “Download Results” button but not appear in the GitLab repository?
Hello, I am having this same issue described above. As you can see in the attached photo, I am in CSV mode. My experiment was previously working as expected, however the data for my most recent 6 or so participants (all collected a few hours ago) is nowhere to be found in associated the git folder.
When I click “download results” I see all the data I expect to find. The “submitted: 11” stat is also as expected.
Why can’t I find them in the git? Is there a way to fix this? It will make my workflow much easier if I can view and organize these files in the git folder as opposed to using the download button.
I just encountered the same issue today as well. I am also in CSV mode. I did a test run with a study script where all the data had saved to both Pavlovia and Gitlab as expected a few weeks ago but now that script is also only saving data to Pavlovia without any copies being saved to Gitlab.
Description of the problem:
We ran 5 subjects through Prolific in the above task just to pilot and make sure everything is working. Everything seemed to go well, except that when I updated the git repository on my computer, there were 7 data files, and 1 participant seemed to be missing. When I check the Pavlovia dashboard it says 5 completed and 3 aborted, and when I download the data from the dashboard, indeed the zip file contains 8 data files. Why then are there only 7 data files in the gitlab repository for the experiment?
I found a similar post from a year ago, Data not saving through Gitlab but there are no responses on possible causes for the discrepancy.
I’ve just amalgamated some threads here and noted that @apitiot commented here that this can happen, but he expected it to be a rare event…possibly caused by participants ending simultaneously. If you are experiencing this issue are you using a recruitment platform like Prolific or some other method which may result in large numbers of participants doing the experiment at the same time. Also, are you using the shelf or an embedded survey at the end of the experiment?
