Hi,
In my survey I have three assessment questions to a number of statements. Each one of these questions has a 4 point likert scale answers plus a fifth Don't know point. E.g. the first question is "How important is this statement to your organization?" and the possible answers are: Not important, Slightly Important, Important, Very Important, and Don't know.
I need help with an appropiate way of coding these response options into numbers that I can use in the analysis as well as with appropriare methods of analysis.
Thanks
Dear Michael,
Place the don't know option in the beginning of four ordered point as,
0:Don't know 1:Not important, 2:Slightly Important, 3:Important, 4:Very Important,or
-1:Don't know 0:Not important, 1:Slightly Important, 2:Important, 3:Very Important.
The respondent having no idea about importance of statement must be placed at the lowest level.
Regards!
Dear Basheer,
Thank you very much for your answer. If I use your second suggession and would like to calculate the mean (even though I understand it is not always recommended), I guess I will need to drop the Don't know answers, ie the -1 answers. In such a case, suggession 2 is better starting from 0 as being not important. What do you think? How would you suggest analyzing these answers by the way?
Thanks again
Michael
If working in SPSS you can (select the option) transform into the same variable, old value 5 (don't know) to new - missing value, and those answers would not be calculated for the arithmetic mean. But before you change the coding, you should check the frequencies, to see how many respondents had chosen that option.
- Badges
- Science method
- Similar topics
- Mathematical Sciences
- Statistics
- Applied Statistics
- Survey Design