Archive for April 7th, 2006

How to analyse answers on open ended questions with SPSS

We got a question from Tom:

“I’m doing research. In this research I had interviews with 20 different
persons. The interviews were open questions (so not only yes/no answers,
but also yes/no/sometimes/only when.. etc)

I am looking for a way to translate the answers into SPSS, so I can make
graphs and percentages for each question (For example: 25% answered yes, 30%
answered no, 2 % answered only when etc.)

Is there an easy way to deal with my problem?”

To begin, SPSS is not meant for analyzing open answers. You can not easily make graphs out of this kind of data. SPSS maximum character input for string fields (used for open answers) is 255. This means you cannot input long pieces of code in the SPSS file.

The only way to create graphs and tables (with counts, percentages etc.) is to categorise the answers the respondents gave you. If you have for example a question with the answers:

Q: Do you you use product 3?


- Yes, I have used it for three years now, mainly for purpose X, Y and Z
- No, I have no idea what product you are talking about. But for purpose Z and B I often use product 4 from this and this supplier….
- Yes, I hate to admit it, but I use it
- Yes, I use it all day, for purpose Z
- Yes, for purpose Z
- No, but would like to use it

You could categorise it into the following variables:

1 = Yes, often
2 = Yes, sometimes
3 = No, but I am interested
4 = No, and not interest to use it

April 7th, 2006 andris


April 2006
« Mar   May »

Posts by Month

Posts by Category