|
|
Hi guys,
I am trying to see which factors increase the risk of having a child with birth defects in a binary logistic regression. I have cases and controls and the variables are dummy. My questions i think are really of a begginer's leve but I can't proceed without askingl:
1- If I have a model with a bunch of variables that are not significant, do i keep these variables on the model? like, if I have age and race and they are not significant, if I want to control for both age and race, should I leave these two variables in the model?
2 - Where do I look to see how much in the terms of percentage the model could successfully explain he variance in the dependent variable?
3 - If I want to show the crude and adjusted odds ratio with confidence intervals, which tool do i use ? I think when we talk about adjusted odds ratio, it means with all the variables in the model,right? If so, where do I find the confidence intervals option when running the logistic regression?
4 - To calculate the crude odss ratio, do I use crosstabs in SPSS? If so, which variables, go on the columns, rows and layer?
Many thanks! Love.
|
Locked
