06-21-2013 02:51 PM
I need to compare means before and after study entry to see if there is a statistically significant difference between the two, and am wondering what the best way would be to evaluate this? I wanted to do logistic regression; however am unsure how to set this up.
I need to do this for three separate models: ER visits, Hospital visits and Jail visits - and need to see if the average visits is different before and after study entry (in order to assess impact of treatment/counseling on these visits).
Here is how the means are setup:
proc means data=CostDrivers;
var ERDAYS_PRE_4 ERDAYS_PRE_3 ERDAYS_3 ERDAYS_4;
var HOSPDAYS_PRE_4 HOSPDAYS_PRE_3 HOSPDAYS_3 HOSPDAYS_4;
var JAILDAYS_PRE_4 JAILDAYS_PRE_3 JAILDAYS_3 JAILDAYS_4;
ERDAYS_PRE_4 = 12 months before study entry whereas ERDAYS_4 = 12 months POST study entry.
06-21-2013 03:07 PM
I think you need to provide some more information regarding the experimental design.
From what you have I don't see why 3 t-tests or 3 anova's wouldn't be appropriate.