Hello,
It was suggested that I use the estat cv command to compare model fit between two logistic regression models (survey weighted). I am wondering whether someone can explain the interpretation of the values in the output. Do you suggest reporting the coefficients or the CV(%) values? Or both?
Code:
Dichotomous outcome. ACESCOREsum is a continuous variable for number of adverse childhood experiences. ace_discr is a dichotomous (y/n) variable for experiences of racial discrimination. Comparing ORs and significance levels is our primary interest, but we thought that looking at model fit/variance might also be useful.
Other suggestions for comparing variance/model fit between the models are also welcome.
Thank you!
Helena
It was suggested that I use the estat cv command to compare model fit between two logistic regression models (survey weighted). I am wondering whether someone can explain the interpretation of the values in the output. Do you suggest reporting the coefficients or the CV(%) values? Or both?
Code:
Code:
svy, subpop(include): logistic MEBhealth c.ACESCOREsum estat cv svy, subpop(include): logistic MEBhealth c.ACESCOREsum ace_discr estat cv
Other suggestions for comparing variance/model fit between the models are also welcome.
Thank you!
Helena
