Correspondence analysis parameters

[etnlessard]

Hello,

I'm still working on my thesis and have a couple of questions concerning correspondence analysis that I have had a hard time finding in books.

My project mainly involves describing the usage of different modes of transportation based on social position, living area etc. Since I wish to do something exploratory, I decided to use correspondence analysis to see how different social factors shape the usage. (i.e. type of education vs. mode of transportation) So far I have 152 online completes but I wish to see how a particular mode of transportation is positioned and of the 152, only 18 use the one I'm interested in.

-Can somebody tell me if my sample size is sufficient or not?
-Also, would it be valid to use a clustering technique on the sociodemos to create some sort of social position indicator and cross them with the modes of transportation?
-Furthermore, is it possible to do mutiple correspondence analysis in SPSS and is this software generally acceptable for CA?
-The multidimensional statistics course I took didn't include this analysis and so I was wondering if you might have any pratical pointers when using this analysis.



Thanks everyone,

etnlessard