Some data shifted four comumns to the right in data file

Builder
1

OS: windows 10 Education
StandalonePsychoPy2-1.90.3-win32

I run an experiment alternating between four different tasks. The variables from one of those taks are consistently saved four columns to the right of where they should be.

One oddity I noticed is that the various counter for this task, .thisRepN, .thisTrialN, .thisN and .thisIndex don’t update at all, anywhere, but in the rows with the task data in question, three other sets of counters for loops that are not active do update. If this displaced the data into other columns, I would expect a displacement by eight columns, not four (data missing in four columns, superfluous data in 12 columns).

Thanks for any comments

Robert Biegler

2

I suspect it will be hard to figure this one out just based on your description. If you could upload an output file and, ideally, your experiment, that would probably make things easier.

Jan

1 Like
3

I worked it out. So, first, I have a rating scale that I customised so that I could stretch it across the screen. The text in the Custom box is:

low=1, high=5, precision=1, textSize=0.35, stretch=2.7,
choices=[“Absolutt Ikke”, “Litt”, “Middels”, “Ganske”, “Svaert”]

I’ll get back to a contradiction in there; PsychoPy doesn’t behave quite as described in the book.

My procedure for getting data into Excel is to convert into text, because when I tell Excel to open a CSV file, it just puts everything into a single cell, without giving me the option to tell it to separate. From text, I first tell Excel that the file will be delimited, then that it will be comma separated. With one exception, that puts everything into its proper place.

The exception is the information from that Custom box, specifically the choices. The first quotation mark around “Absolut ikke” is lost, and then the remaining four choices are saved in the next four columns, which are supposed to be reserved for other data. And that is what shifts everything that is further to the right by those four columns.

If I knew how to tell PsychoPy how not to save data, I could just tll it to omit this stuff in the CSV file, and then nothing would get shifted. I don’t know whether such a command exists. I have failed to find one.

As for the contradiction in the Custom Box, I want a numerical rating, hence the low and high values. Specifying categorical choices interferes with that. However, when I say these are labels, then, contrary to what is in the book, I don’t get my five labels evenly distributed under the scale. I only achieve that by asking for choices. Then I have a bit of code which converts them into numerical ratings.