s********8
发帖数: 50 | 1 there are more than 5 variables in the model: A.B.C.D.E.
A. B. are not significant (p>0.05) when they are in the model separately or
in the model togather. But when I put A|B (=A,B,A*B) in the model. Both A(p=
0.01) and B(p=0.02) are significant , but interaction term is not
significant (p=0.08).
what does it mean for modeling ?
do I need put three of them into final model?
Thanks |
D*********2
发帖数: 535 | 2 interesting.
What about A alone, I mean, not even C,D,E. Just A. |
r*****y
发帖数: 199 | 3
I think he/she already mentioned that A alone is insignificant, B alone is
insignificant.If I were him/her I would choose A B and A*B all in the model.
【在 D*********2 的大作中提到】
: interesting.
: What about A alone, I mean, not even C,D,E. Just A.
|
r*****y
发帖数: 199 | 4 If you are doing a model selection, I think you can try both stepwise or
backward. I personally prefer backward. |
P****D
发帖数: 11146 | 5 What are you trying to do? Build a prediction model, or analyze the
contribution of ABCDE? |
l*********s
发帖数: 5409 | 6 put all of them into the model |
s*****9
发帖数: 108 | 7 .08还不错啊，可以放进去的。要是interaction term在你的model里面make sense的话
就考虑进去吧 |
s********8
发帖数: 50 | 8 what;s the difference between build a prediction model and analyze the
contribution of ABCDE?
【在 P****D 的大作中提到】
: What are you trying to do? Build a prediction model, or analyze the
: contribution of ABCDE?
|
s********8
发帖数: 50 | 9 I tried stepwise, backward and forward. and then I picked variables in
common in the model first. Next I tried to add variables accepted in only
one method or two mehtods in to the previous model.
【在 r*****y 的大作中提到】
: If you are doing a model selection, I think you can try both stepwise or
: backward. I personally prefer backward.
|
o*p
发帖数: 177 | 10 如果A B是你的design variable，那就要放在model里面别动了。比如你的研究想看男
女有没有区别，或者黑白有没有区别，那么gender race都放在model不要动了。
如果只是找predictor，那A和B都可以不要了。
【在 s********8 的大作中提到】
: what;s the difference between build a prediction model and analyze the
: contribution of ABCDE?
|
