Basically I would like to understand that when I am doing analysis as follows
Selected Analyze > Reliability and Survival > Fit Proportional Hazards. Selected Time as Time to Event, Selected censor as Censor, after this I transferred my variable in construct Model effects using "Add" At first I run this keeping one variable in construct Model effects and after this I "run" this keeping all three co-variate together. Now could you please conform me that the out come of first analysis (using one variable) is for univariate and 2nd analysis ( keeping all three co-variate together) are multivariate analysis.
I think that by definition the first set of models is 'univariate,' though I have never heard it referred to that way. I do not see the value in such an analysis. The second way is definitely 'multivariate.'
Ok, Could you please suggest me that, How can I do univariate analysis to see impact of one of my continuous variable(gene expression) upon survival time in my cohort?
I am confused. I thought that you said that you modeled the response separately with each predictor and then with all the predictors.Do you understand the Cox proportional hazards model and how to use the regression results?