I have three items ( safety related questions measured in ordinal scale 1 to 5) and after factor analysis they result to one factor. I have got the factor loading in spss. How can I incorporate this result in multinomial logit choice model. Please help anyone.
Hello Tarapada,
There are several ways to accomplish this. The easiest way is to rerun the factoring, and have spss save factor scores to your data set. In the Factor dialog box, click on "scores" button, then check the "save as variables" box.
A more common method is to simply sum the scores of the three items (assuming that loadings had the same sign) into a total, and use the summated score as your result.
Good luck with your work.
Hi Tarapada,
Although you have an answer to your question already, I would like to point out something that may help you in the future. As far as I know, factor analysis is not a suitable solution when you have categorical data.
The problem is the following. In the case of questions measured in ordinal scale, 1 to 5, a change from 0 (very bad) to 1 (less bad) is identical to a change from 3 (regular) to 4 (good), 1 unit. Unless you can prove that all the people in your dataset value those changes equally, you are introducing a bias. Is the improvement from regular (3) to good (4) valued the same that the improvement from good (4) to excellent (5)?
A solution to this problem is to use another kind of analysis. In your case, an option would be to create as many dummy variables as values you have in each variable, 5 dummies per variable. Nonetheless, that may create a problem of over-parametrization. In this situation, if you want to reduce the dimensions of your dataset, a good option is the factorial correspondence analysis, that intuitively, is the "factor analysis" of categorical data. If this is the kind of factor analysis that you are carrying out, sorry for the long post. But I think it is worth highlighting that there is a difference between both factor analyses.
Best regards,
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs