Pavlovia: Session file variables not transferred to log file for last participants (first observed Sept, 2, 2019)

Online experiments
21

Hey Jon,

I was using the same experiment link for 2 months with no new updates- https://run.pavlovia.org/sparmar34/group_3/html/ . I did update the Psychopy version on my desktop but I never synced my previous experiments again from builder to Psychopy or vice-versa. All the data collected in last 12 days does not have any trial information. Is there a way to recover that somehow? I was using experts for my Study and I know they are not going to participate again for an hour. Any help would be really appreciated.

Thanks.

1 Like
22

Hi Joh,
I am in a similar situation and will be interested in knowing if there is any way to recover these data.
Thank you,

23

Firstly, thanks for all the provided information. As @Sweta_Parmar, I would be interested if there is any data on the server with which the collected (but in the first half of September wrongly saved) data could be restored?

@jon, as I at the moment also face issues with my older code (from 3.1.5) on newer versions (3.2.X) independently of Pavlovia, version (b) of your descriptions sounds as the better solution to me!

Furthermore, I have a suggestion for the piloting status. It would be a great help of being able to collet 1 to 3 complete data sets in the piloting status to check if the data looks like it should look like. I seem to understand why that is not the case at the moment, but on the longer run a small amount of “piloting trials” with data collection would be an amazing feature. Probably it would also work without that. I just could imagine, that it would convince more people to afterwards buying the credits.

24

No, unfortunately we really can’t retrieve data from studies that suffered from this loss. Very sorry.

Yes, we have gone with option b) of fixing the version of the PsychoJS lib to the version of the PsychoPy app that compiled it. You can still use the Settings>useVersion setting to customise exactly what version is used in the compilation but avoid versions 3.2.0 to 3.2.2 because you’ll be int he same boat as before!)

Further measures we’ve taken to prevent any further loss:

  1. We’ve added further code so that the conditions data are captured even when the lib/app were not correctly paired as above. Right now, any further data collection in your studies won’t be suffering this issue even if you haven’t re-synced your code
  2. We’re working on adding log files right now, similar to those in PsychoPy itself, so that there is a second record of what happened in the experiment (Belt and braces to keep the trousers up!). In this instance a log file would have allowed us to fix the csv file but it hadn’t yet been programmed to output.

best wishes and our apologies again,
Jon

25

sorry, what does “modify the head” of your experiment.js file means, I can see the file under the html folder, but what is the “head”?

26

I believe it’s the beginning of the file, as described by @apitiot in this thread.