s*********e 发帖数: 1051 | 1 it is very interesting http://statcompute.wordpress.com/2012/09/30/marginal-effects-on-binary-outcome/.
you might test it yourself. as shown in many research literature, marginal
effects across different classes of models can be very much similar even
with substantial differences in model specification and assumptions. this is
very useful information in the content of model comparison. |
A*******s 发帖数: 3942 | 2 非常有趣,读lz的blog受益良多。
logit和probit实质上没有什么差别,因为standard normal distribution和scaled
logistic distribution with unit variance的形状是很相近的。logit和probit的系
数的关系大概是1.6至1.8的关系,但是对应的std dev也是同样的倍数关系(因为
standard normal和standard logit的square root of variance也是同样的倍数关系)
,所以两者的t statistics和p-value非常接近。marginal effect是相同应该也是同样
道理吧。
其他model为啥marginal effect相同我就不清楚原因了。
is
【在 s*********e 的大作中提到】 : it is very interesting http://statcompute.wordpress.com/2012/09/30/marginal-effects-on-binary-outcome/. : you might test it yourself. as shown in many research literature, marginal : effects across different classes of models can be very much similar even : with substantial differences in model specification and assumptions. this is : very useful information in the content of model comparison.
|
s*********e 发帖数: 1051 | |
s*********e 发帖数: 1051 | |