Reducing Number of Categorical Variables

I have an abundant amount of categorical variables pertaining to an epidemiological study. I need to filter these down to a succinct list and wish to develop an overall model as well as sub-models based on a specific environmental categorical attribute as an independent/ explanatory variable. i.e. DV 1,2,3 are more likely with IV model A and 4,5,6 are more likely with IV model B

I have looked at Categorical Principal Components Analysis (CATPCA) and Discriminate Analysis but am not totally sure this will provide the results I require. Does anyone have any suggestions as to how I can reduce the number of categorical variables I have?


No cake for spunky
There are a couple of possible approaches depending on what you want to do. One is to look for underlying dimensions behind your IV through Factor analysis and then analyze the factors rather than the variables. Another is to limit the variables through what research suggests is important (or past findings). If you have data for these variables already you could run regression (or ANOVA) and remove variables that are not statistically signficant.