w*********m
发帖数: 196 | 1 When doing a regression of Y and X1, the coefficient is positive significant
When doing a regression of Y and X2, the coefficient is positive significant
When doing a regression of Y and X1 and X2, the coefficient of X1 turns out
to be negative significant,and the coefficient of X2 remains positive significant.
What might be the reasons?
Is is because of collinearity?
Thanks:) | s*r
发帖数: 2757 | 2 looks like a version of simpson's paradox | w*********m
发帖数: 196 | 3 I got it. Thanks:)
Will there be other possibilities? | w*********m
发帖数: 196 | 4 I searched the website, it is more about ratio.
I am wondering how can we apply it to regression? | r*****e
发帖数: 71 | | T******r
发帖数: 265 | 6 1. You can calculate VIF to see how bad collinearity is, or a simple
correlation 'cause you only have two variables.
2. Maybe create the ratio and run
y ~ x1 + x2 +x1/x2 (or x2/x1)
【在 w*********m 的大作中提到】
: I searched the website, it is more about ratio.
: I am wondering how can we apply it to regression?
| s*****n
发帖数: 2174 | 7 这个太正常了, 比如你随便构造一个 x2,
和一个与x2非常正相关的y, 就满足了第二条.
然后构造x1 = x2 - residual(y ~ x2)
这样满足以下两点.
1) x1 和 x2 很接近, 如果只用y来回归, 都是正相关的.
2) x1 和 对 y对x2回归的residual 是负相关的. 也就是当引入x2以后, x1系数就是负
的了.
本质问题, 就是 x1 和 x2 并不是 不相关的.
significant
significant
out
significant.
【在 w*********m 的大作中提到】
: When doing a regression of Y and X1, the coefficient is positive significant
: When doing a regression of Y and X2, the coefficient is positive significant
: When doing a regression of Y and X1 and X2, the coefficient of X1 turns out
: to be negative significant,and the coefficient of X2 remains positive significant.
: What might be the reasons?
: Is is because of collinearity?
: Thanks:)
|
|
