Logging Trouble
Hey again all,
me and my colleagues recently published an online experiment using a OSWeb experiment as the main component (and only component where the logging is supposed to be not "none") on mindprobe. But we face some weird troubles with the result data, which we did not see while testing.
One first issue seems to be that for a lot of experiments only "DATA RETRIVED" is logged as state. For some of such cases the result data of our main (and final) component indicates that the participants did indeed finish it. We think so because the last row of our logging documents a questionnaire inside the OSWeb experiment, saying this being answered indicated that participants did in fact finish it, and spent ~30 minutes to complete the whole experiment. This leads to two questions.
First, does anyone of you have an (technical) explaination for this? It is possible that participants did just not continue from out finishing screen where we tell them to press any key to complete the experiment, but for those cases I described, which seem to have put in the time and effort to complete the experiment (and which are getting paid for it), this seems rather unlikely (but sure still possible).
The second and maybe even more confusing problem we face for such cases is that even when we have their (almost) complete data, it happens that the result json (data.txt) is not correctly formatted. Often we see e.g. that a linebreak is pasted in between all rows of our loggings, so that the structure is:
{"data" : [{...},
{...},
{...}, ....}]
In other cases the "context" data is no logged and the structure of the data ends aprupt like:
[{...}, {...}, ...,{...},
Did anyone of you face similar problems before? Would you imagine this being rather JATOS or OSWeb related?
Thanks a lot in advance,
Leo
Comments
Hi Leo,
Sorry for the slow response, it’s holiday time already :)
I don’t fully understand your questions. For the first one, you seem to have an unlikely but plausible explanation. But, if I understood it correctly, it doesn’t make sense to me. The results of the first component must be sent to the database (to the server) before the participant can continue to the next component. So I can’t imagine what’s going on. Would you mind describing your problem again with some more detail? Which data exactly should be, but aren’t, getting saved? And how exactly is that questionnaire setup? Is it a separate JATOS component, or a separate part of the same script?
it would be worth checking the underlying code, how you (or OSWeb) is passing the data and starting the next component.
The second question is, I think, more for OSweb. As they are the ones sending data. It sounds to me that the study was interrupted, or the participant closed the tab before the data got sent. But I don’t know for sure and it does sound like a front end thing.
But please do explain again if you think these answers don’t make sense - maybe I just misunderstood.
Best