G**7
发帖数: 391 | 1 Anyone having knowledge about weighting complex survey?
Due to various reason, some software or skills (e.g. random forest) have not
incorporated the weighting yet.
My question is: can we proceed to multiply all the independent variables of
interest by the weight and proceed to get the result?
Not sure.
Thanks. | f***h
发帖数: 283 | 2 No. you can't. That's not the way it works. Try a simple model, compare the
results from the method you mentioned and from a correct method. you should
see differences. |
|
