# combining two categorical variables

I have one five point Likert scale variable (importance levels) for accessibility to a certain facility, and another three-level categorical variable (preferred distance).

I want to combine these two variables into one variable so I could relate it to the current distance of that facility.

How do I combine these two variables? Should I use cluster analysis?

I have edited the question as per Nick's suggestion:

Hi, The preferred distance is categorical-0.5km, 2km, 5km, I numbered them as 1,2,3 respectively. I did relate them separately in bivariate Chi-square test. But I believe, there is a relationship between importance level and preferred distance. For ex: medium importance may be opted with medium distance. If I combine these two, I could say that there these two variables act together in making the decision. thanks.

• Is preferred distance ordinal or nominal? (Different answers will apply in either case.) What do you think this one variable would really mean? (The meaning may be clearer without combining them.) Are any other indicators of it available? (Two isn't really enough for latent variable estimation.) Is there a reason you'd be less interested in bivariate relationships among these three separate variables? (This would be much easier, and probably much clearer.) – Nick Stauner Apr 3 '14 at 1:16
• Hi, The preferred distance is categorical-0.5km, 2km, 5km, I numbered them as 1,2,3 respectively. I did relate them separately in bivariate Chi-square test. But I believe, there is a relationship between importance level and preferred distance. For ex: medium importance may be opted with medium distance. If I combine these two, I could say that there these two variables act together in making the decision. thanks. – aruna r Apr 3 '14 at 20:53
• If you edit the appropriate information into your question, that will bump it up to the top of the active questions list and get more people to take a fresh look at it. – Nick Stauner Apr 3 '14 at 20:56