Hi all,
first, thank you for such a great mailing list. I worked with a lot of
software and never had such a helpful mailing list. :)
I ran into a problem and I found two possible solution and I am not sure
which one would be the best.
My experiment consists of six blocks with a consequent memory test per
session per participant. I want to run an analysis that looks for
activation for correct remembered pictures and incorrect/not remembered
pictures. I also want to run some contrast on these two EVs.
Now, it happens of course, that for some blocks, people have no correct
or incorrect answers. So, the corresponding timing file would be empty,
crashes would occur and I would be very sad. Especially, as I want to
combine the session blocks in a higher level and I need the copes to be
there for every block.
I searched the mailing list and found two ways of dealing with it.
First, I could replace the EV with the empty timing file with an empty
EV. (Message-id:
<[log in to unmask]>)
Another way might be to replace the empty timing file with a de-weighted
entry such as 0 0 0. (<[log in to unmask]>)
Which of these approaches would be "cleaner"? Are there other ways to
deal with this issue?
Thank you so much for your help,
Andrej
|