One of my questionnary made with Google Doc have a multi response set.
Let's say the three possible questions are :
- foo
- bar
- foobar
People where allowed to tick one or more answers.
In my google spreadsheet results, I have :
answer1 : foo
answer2 : foo, bar
answer3 : bar, foobar
answer4 : foo, foobar
anwser5 : foobar
etc.
How could I analyze this type of result in SPSS ?
I have the intuition that the values could be split in multiple answers to multiple questions by I don't know how to do this.
Thanks
You could also try to use the index function when creating a new variable in SPSS.
For example if the answers to be processed are in the ANSWER variable, you can choose Transform / Compute variable..., and make a new numeric variable named FOO, using this function:
CHAR.INDEX(UPCASE(ANSWER),'FOO') > 0
This will result a 1 if foo is found in the ANSWER variable, and 0 if not. Then do the same with the other possible answers.
Or you can simply run this syntax:
I do not have SPSS to test it, but this solution could help you http://www.spsstools.net/Syntax/MultipleResp/ParseCommaSeparatedLetters.txt
To split this string into three variables:
Create the three variables, copy the strings, paste in Excel and use the special text option, choose delimiter = comma. This will split the text of the string into three Excel cells. Then you copy and paste into SPSS, all three variables at once.
Dealing with text/data, SPSS splits by variables whenever it meets tabulation (paste into Excel adds the tabulation character where there was a comma). Alternatively plain text editors (free: TextPad, Notepad++) can be used to alter the text.