psychopy.org | Reference | Downloads | Github

Credit consumption and saving (incomplete) data faulty

URL of experiment: https://run.pavlovia.org/Rika/masterarbeit-bevi_studie2021/html

This topic is not new but I can’t seem to find any good solution for this. Maybe someone could take a look at my experiment again and see, if there is something going wrong? Possibly this could also be related to my former topic from two months prior: Unspecified JavaScript error & error closing session - #25 by Rika

I started running my experiement a month ago and since then I do not have the feeling, that all data is saved correctly. Since the beginning 82 sessions ahve been started but the rest of the number do not add up (Credits consumed/ aborted sessions; see picture below). A lot of times a credit is reserved, but then just expires after 24h without any data or log info being saved. Although I do have the “save incomplete results” switched to on (and csv data saving format). I know for a fact that some participants started the experiment and for whatever reason aborted the experiement, yet I cannot see any results. Now I am wondering, how many times a participant actually completed the experiment but the results were not saved (not even the incomplete ones)?

Then yesterday the experiement was started 3 times and I received 3 csv files. One empty (1KB), 2 with incomplete results. But only one credit has been consumed and the other two are still reserved (and will expire in a few hours)… I don’t understand the system here? Shouldn’t have been two credits consumed since I do allow for incomplete data to be saved (although this worked in my favor)?

One participant also received an error message when trying to end the experiemnt and save their data.

Also sometimes there is a gz file with a csv file in my repository on gitlab, and sometimes not. And this is independant of whether the results are complete or not.

I’m losing my mind here. Never thought running an online study could be sooo nerve wracking :sweat_smile: :joy:

Hey again,
K, let’s check it out then. Could you share your repo with me? (tpronk)

1 Like

Hi again :smiley:
Thank you! Added you to the repo as maintainer :slight_smile:

I took a look at your psyexp file and I’ve got two (general) recommendations/possible explanations.

  • About the experiment. I notice quite a lot of funky code components in there, with some creating objects in Each Frame tabs. This can have two risks: (1) it might be that some of the code isn’t supported on all browsers, (2) if the creation of objects at each frame has some issue, this can cause “memory leaks”, i.e. the memory gets filled up an the browser crashes. I’m not sure if this happens for sure, because the code is too complex for me to analyze in reasonable time period, but I’ve seen such issues often in the past.
  • About the data. When the browser crashes, but you’ve got “save incomplete data” enabled, Pavlovia does consume a credit while the browser hasn’t been able to submit any data yet. That might explain the empty data files.