Cluster Analysis

Hi all,

I have a small sample of 75 participants. I'm conducting an exploratory study looking at psychological distress and gender role variables...thus, I really have a lot of variables (15 or so) that I want to look at, but the sample size is so cluster analysis the best way to go? I thought a canonical correlation may be efective, but I'm guessing I would need to create many composite variables (plus I was told I would have trouble publishing because can. cor. is "outdated"). I was also told to pick one DV and run a multiple regression. Any help would be appreciated!