| Reference | Downloads | Github

AddOtherData does not add data to output in QUEST staircase

I’m trying to run QUEST staircases in Psychopy 3.2.4.

My staircases seem to be running fine, however I can’t seem to add additional data to the output file. I want to add some extra parameters indicating particular characteristics of the stimuli - these aren’t part of the staircase because they are randomised from trial to trial.

My extra code snippets look like this:

trials.addOtherData('resp.keys', str(resp.keys))
trials.addOtherData('resp.rt', resp.rt)
trials.addOtherData('correctAns', correctAns)

But no additional data turn up in the output files. But my code trials.addResponse(resp.corr) works. How to solve this problem?


What function are you using to export the data? I just looked in the source code and upon a cursory inspection, it seems that “otherData” is only exported with .saveAsExcel() and not .saveAsText().

Could you try testing with both and let us know if there is a difference?

PS: I’m not sure if this is a Coder or Builder question. If Builder, you select the output formats from the experiment settings dialog box.

I used:


I tried to replace the following code with .saveAsExcel(filename+'.xlsx') or .saveAsText(filename+'.txt') . But the output is still the same. I mean I still cannot get the output I want and the data file I got is still csv file.

By the way, I am using MultiStairHandler now to do the QUEST staircase experiment without interleaved trials. I wonder whether it is better for me to use QuestHandler to do the QUEST staircase experiment?

Hi Amelia, instead of thisExp.saveAsExcel(), what happens if you try to save directly from the staircase itself, with trials.saveAsExcel()?

It showed
fileCollisionMethod=fileCollisionMethod, encoding=encoding)
TypeError: openOutputFile() got an unexpected keyword argument ‘delim’

Oh wow, you really are opening up some cobwebs in the code here… On googling, a few other people over the years have had this issue with .addOtherData() not having any effect either, so it looks like something that really needs to be addressed.

Are you able to post a minimal experiment here which shows the issues? i.e. one with ideally no images or other resources, and that just runs for a few trials? That would help with testing and debugging the issue.