| Reference | Downloads | Github

Data not saving if internet connection is poor at the end

URL of the experiment: day5_group2_test_training_test [PsychoPy]

Description of the problem:
The problem is not specific to this experiment.
AT the end of the experiment, Sometimes data is not being saved and the participant is getting this message:

The following details might be helpful for understanding the issue better:

  • I have clearly defined the end to the study. Participants are instructed to press “space bar” to end the task.
  • Participants are instructed not to close the screen until data saved confirmation comes (which is after few seconds)
  • This was a 7 days study with two groups of participants. Hence, 14 experiments were running. 6 days experiment was 30 minutes long, 1 day it was 1 hour 15 minutes long. Only once saving failed in half an hour session. 5 times it failed in the longer session. The longer session was on various days for different participants. The same participants have already done 30 minutes sessions so they do follow the instructions.
  • “saving incomplete data” was ON, so even if they would have terminated the experiment, I would have got incomplete data (I guarantee it’s not a problem from the participant’s side, they are regular participants in our lab).
  • Other members of our lab are also facing this issue. One member has already mentioned the issue here.

What I understood is, this is happening if the internet connection is not steady at the end. The server tries to upload the data, and if it can’t, then the data is lost. For longer experiment, the data size is larger, hence, the chance of data loss is also more.

We will start a project on children to see their development in the task with age, so will collect data from each child at regular interval. It will come very costly if we lost data there.

Please help me with this issue. Can there be a way so that even if there is complete internet loss, the data will not be lost? At least that will be downloaded at the local computer? That will be great!

Jhilik Das.
Ph.D. student.
Vision Lab, Indian Institute of Science.

Hi There,

Unfortunately this is not something that we are able to control. I would recommend encouraging your participants to only start the experiment if they have a stable internet connection.


Hi Becca,
Adding an option to download the data remotely by the participants will really solve the problem. Probably on the experimenter side, you can add an option like “allow participants to download data”?
Will it be too difficult?
It will be really helpful for me (and I am sure for many others).