View Full Version : what test to do?

05-23-2009, 03:51 AM
Would be grateful for a little guidance.
I have a set of data with a mix of continous and categorical data

I am trying to see if there is any association between a continous variable ( lipid level) and 3 categorical data, smoking, gender and marital status (this has 4 categories i.e single, married, divorced, widowed). I was thinking of using cross tabs but guess that would mean trying to categorise the lipids into low normal and high.
Any suggestions welcome.

05-24-2009, 04:05 PM
I think you should try running a linear regression model ( if the assumptions are correct ), and make dummy variables out of the IV you have.