### spss >> Entering multiple response data to 1 question

In SPSS 11.5 for windows I have a survey question that reads- please
check all that apply - neck back elbow shoulder knee wrist
other

WE are trying to figure out how to enter this data into SPSS - we have
no problem if the person has 1 problem (entering as string data)- but
how do we get SPSS to accept data if the person has more than 1
problem (eg wrist and back)

Do we have to seperate the question and do strings for problem 1
problem 2 etc. Or is there an easier way to handle this data.

We are entering the data for approx 80 questionnaires.

Many thanks
Steve

### spss >> Entering multiple response data to 1 question

You need to spearate out the question so that you have a variable for
each option, eg you would have two variable wrist and back, which
would be yes or no questions, ie the person either has the problem or
not.

If you need to combine the answers in analysis you can use multiple
frequencies

Fred

### spss >> Entering multiple response data to 1 question

You need to spearate out the question so that you have a variable for
each option, eg you would have two variable wrist and back, which
would be yes or no questions, ie the person either has the problem or
not.

If you need to combine the answers in analysis you can use the
multiple resposne option

Fred

### spss >> Entering multiple response data to 1 question

You need to separate out the question, so that every response becomes
it's own variable, eg you would have two variables neck and wrist,
both of which would have the option yes or no, ie they either have the
problem or they don't.

If you need to combine the responses in analysis you can always use
the multiple response option

Fred

### spss >> Entering multiple response data to 1 question

Usually data of this kind are entered as a number of dichotomous variables,
each indicating whether the alternative was checked or not:
neck: no/yes -> 0 or 1
back: no/yes -> 0 or 1
etc.
Someone who has checked wrist and back would then be entered as:
010001

The "other" category poses another problem. You could enter the response to
this question as a string, but since you have only 80 questionnaires, perhaps
it's possible to skim through the questionnaires first and see if there are
responses/categories that occur rather frequently, e.g. "head", "feet". If
that's so, you might consider extending the number of dichotomous variables
with these categories.
Once the data are entered, use the MULT RESPONSE-procedure for a first global