Loss of variables in data

Elliot_Luo

Hi,
I ran into a severe problem recently and any advice and suggestions will be greatly appreciated...
My experiment contains two different but related tasks, and subjects had to complete both. To counterbalance, I put both of task (let's call them task A and B.) inside a loop, and used inlines and run-if of sequence inside the loop to control the order of the tasks.
When subject_parity is odd, the order is A then B; otherwise, the order is B then A
All the variables were imported using different .csv file.
The experiment ran successfully, but all the data of odd subject_parity only had variables in task A, and the data of even subject_parity only contained variables in task B.
I used a single logger in each task to log data, and Log all variables was enabled.

I wondered whether logger would only log the first encountered variables, i.e., the first external file?
If so, is it possible to log all the variables?

Thanks in advance,
Elliot.

sebastiaan

Hi Elliot,

What's happening here is the following: OpenSesame doesn't know in advance which variables are defined in your .csv file, as it does for example for the loop table. So when the logger is executed for the first time, and the columns in the logfile are created, only the columns that OpenSesame knows about are included. If you then, in the second task, add more columns by loading another .csv file, these won't be included. Does that make sense?

To work around this, you could do either of two things:

• Add empty placeholder columns to both .csv files, so that they both contain all columns; or
• Explicitly add the missing variables to the logger.

Cheers,
Sebastiaan

Elliot_Luo

Thanks Sebastiaan!
I would take your advice to revise my experiment