|
|
We just finished running a study in our lab, and we've noticed some strange data shifts. For some of the participants, the data is split for some reason. It appears on several rows instead of being organized in one row. It only occurred in a minority of the cases, but it is concerning because sometimes there is missing data where the break in rows occurs. I have never seen this happen before. Any ideas as to how to prevent this from happening in the future?
That can happen if certain characters are included in the response data. Usually it's something like a line feed, hard return, comma or something that makes Excel, SPSS or other analysis program think that the current input line has ended. What version of MediaLab is it that you are using (i.e., 20xx.x.x)? And can you identify the item type (e.g., fill-in-the-blank, custom item) at which the shift seems to occur?
Thanks for your help!
We're using version 2008.1 and this shift happens at a fill-in-the-blank item.
I will try to experiment with ML to see if any of the characters you've mentioned is responsible for this issue.
Great. Another thing that might help is to open the data files (csv and/or txt) in WordPad or NotePad and see if you can see any commonalities in those cases where the line breaks occur.
