Hi, just one short remark:
from my decade's experience in opinion and market research:
You should never use this precoded and predelected "no answer". People like comfortable things and what is more comfortable than this preselected option to proceed in the survey without having to use your brains?
In most cases, really in most cases you can define your own "DKNA" code with the advantage that in the analysis, let's say by SPSS you have ONE "missing value" code.
And you can style it that this answer option is less suggestive (smaller font, grayed, or so)
tbart said: using bipolar matrix questions (those defined by subquestion text "not at all | very much"), it is not.
In this case I wouldn't use the pipe symbol, but have these option in the header of the matrix.
Best regards
Joffm
from my decade's experience in opinion and market research:
You should never use this precoded and predelected "no answer". People like comfortable things and what is more comfortable than this preselected option to proceed in the survey without having to use your brains?
In most cases, really in most cases you can define your own "DKNA" code with the advantage that in the analysis, let's say by SPSS you have ONE "missing value" code.
And you can style it that this answer option is less suggestive (smaller font, grayed, or so)
tbart said: using bipolar matrix questions (those defined by subquestion text "not at all | very much"), it is not.
In this case I wouldn't use the pipe symbol, but have these option in the header of the matrix.
Best regards
Joffm