这是问了个什么意思

问题如下：

What are the costs and benefits of dropping highly collinear variables?

选项：

解释：

The cost is that the model may be misspecified if the dropped variable should be in the true model. The benefit is that the parameter estimates of the remaining variables will improve because the OLS estimator will be able to clearly attribute movement in the left-hand-side variable to the remaining variables.