problems with coroutines on osweb in logging data (null/no responses)
Hello,
I am a new user to opensesame/osweb. I implemented my experiment in open sesame and i use coroutines to log responses. i then exported it to osweb and running it on jatos. However, when i look at the log files, the trials where i did not make any responses has the last response (either a 1 or 2 which are the allowed responses) instead of a 'none' or 'null'. This works perfectly when i run it locally on open sesame. Any solution for this? Or alternatively if there is a way around coroutines, that can be implemented in a osweb friendly way? I m using coroutines because i want to show an image using sketchpad for 1000ms, which is followed by a blank screen for 400ms and i want the subject to be able to respond anytime between 0 and 1400ms from the start of the trial. Is this possible to implement? Any insights would be greatly appreciated.
Comments
Hi @vikramrao ,
coroutinesare not currently supported by OSWeb. That being said, @Daniel has been working on implementing them, and you've been using this as-of-yet unfinished (and buggy) implementation. But maybe @Daniel would like to know more about the issues that you're running into, so that we can polish the OSWeb implementation ofcoroutinesfor eventual stable release. For now, it would be helpful if you attach your experiment, or a version of it that illustrates the issue, to this discussion.Cheers!
Sebastiaan
Check out SigmundAI.eu for our OpenSesame AI assistant!
Hi @vikramrao ,
Thanks for notifying us about this. I noticed this myself too. The response data of the previous trial transfers to the next one if no response has been given, and this only happens when using coroutines.
I haven't had the time to look into this to see what causes this behavior, but I bet (/hope) this is a small thing to fix. As a temporary workaround for this, I explicitly (re)set the response vars at the start of each trial using an inline_javascript item that I placed at the top of my trial_sequence, containing:
I hope this helps you for now too.
Thank you so much @Daniel and @sebastiaan. This resolution worked. Explicitly Nulling the variables of interest before every trial worked.
No problem! Glad it worked this way for now!