How to analyze this data

reddy

New Member
#1
Hello,

I need guidance in analyzing the outcome.

I have 4 independent variables, and 2 dependent variables. The sample size is ~600.

I have two groups: Treatment A and Treatment B. I want to determine if treatment B is better than A in achieving superior outcomes.

The independent variables in each group are #1: Continuous, #2: Ordinal (3 levels), #3:Ordinal (5 levels), #4 Categorical (Yes/No).

The dependent variables are #1: Ordinal (3 levels) ; #2: Categorical (Yes/No)

The dependent variable #1 & #2 is based on the decisions factoring all independent variables.

The dependent variables #1 and #2 are competing goals, i.e., to achieve one, the other sometimes suffers. That is, they are dependent on each other (especially in high-risk cases). In low risk disease, both outcomes are achievable without compromising one another.

I want to know how best to address analyzing this data: What test(s) to use to determine which independent variables have a significant effect on each or both of the dependent variables; Which treatment achieves optimum goals in both the dependent variables. And, How to determine the effect of of one dependent variable on the other.

Thank you
 

rogojel

TS Contributor
#2
hi,
the second dependent variable cam be clearly analysed using binary logistic regression. Ordinal logistic regression is also a possibility, for the first outcome, but maybe you can simplify your life by turning the 3 levels into two? E.g. by turnig the levels good, neutral, bad into good, not good?

Once you have a model you can use several tools to find an optimum, e.g. Solver in Excel or Optimizer if you work with Minitab.

regards