Suppose you have delete several observations that had small residual values. If you re-estimate the regression using this reduced sample, what would happen to "standard error of estimates(SEE)" and “R-Square"?
答案指出：删除了observation that had small residual value后，会degrade strength of regression. 所以导致SEE上升，R-Square下降。请问为什么？？我理解的答案意思是当删除了那些小residual的observation后，就只剩下一些大observation了，所以regression的解释程度就差了。请问答案是这个思路么？？谢谢！！
[此贴子已经被作者于2011-5-9 3:28:03编辑过] | 
发表于 2011-5-5 22:30
| 