Good afternoon everyone,

I did a survey and now I’m in trouble. My problem is the following: when I export columns to Excel, I find on my worksheet as many columns as the alternatives were in my research (i.e. when asking “how often do you…?”, I have 4 columns from the same item: never-some times-often-always, representing the 4 alternatives we gave to the subjects). The point is that there is a “no” (or a 2 if converted), under each column till the respondent chose the answer, where I find a “yes” (or a 1) (i.e. a no under the “never” column, a no under the “some times”, a yes under the “often” column, and a no under the “always” column, if the subject meant to say “often”).

I really need to have just one column instead many, with only one number under it, representing the alternative chosen from the respondent (i.e. under the column “how often do you…?”, to see just a “3” or other numbers, instead of a “no-no-yes-no”).

Is there any method to export a single column per answer, now that the survey ended?
If no, is there anybody that knows a method to merge the columns on excel without loosing any precious information?

Thank you in advance

This only happens when you use the question type of multiple responses. In this case it is important to have more than one column, because a respondent could answer more than one subquestion with "yes". In the case of your question I don't know why you chose this question type, because your question would fit a normal array question (only one answer per subquestion possible).

You will have to create your own column correcting that in Excel. There is no way to do this in Limesurvey.

What I don't get is why you chose this question type for this question in the first place. You need to check if someone has responded "yes" for more than one answer option.

Thank you holch,

We chose this kind of items because we wanted the group of questions to look like a semantic differential scale in order to maintain the original layout (the items were from a previous research, made by other researchers) and yes, the subjects were supposed to choose only one answer among some alternatives. However, now the survey is over and we're trying to figure out a method to fix this mistake, otherwise we must do it manually.

In any case, it means a lot to me to know that I can't do it with limesurvey, so I can ask somewhere else.

thank you again for your precious advices

I don't know why a semantic diferential would need a multipe response question, but I guess this is done now...

It shouldn't be too complicated to be able to find a formula in Excel to read the previous columns and then write a code for the column that contains "yes".

For example: Create a copy of the columns in question, just to make sure you can always go back and to destroy the whole data set.

Then subtitute the "yes" with the required code and delete everything else from the column (all "no" answers, etc.). Then create an extra column that reads the previous columns that should now have a code (1,2,3,4, one code per column) and then you should have one column and with the codes. Should take you about 5-10 minutes in Excel (if you know what you are doing, of course).