--- In Statisticians_group@..., Bogarth Hernández
<crisanty@...> wrote:
>
> I have a question about what do you understand as "points of
influence"
> in a Multiple Linear Regression? and how to do tests about it?
>
> I'm sorry about my bad English.
>Dear all,
points of influence are those that are far from the spce of the data
which means that they are outliers or leverage points or both.
remember that the traditional technique assumes that the data is
homogeneous and this assumption isnot satisfied in case of points of
influnce.many tests are used to discover points of influence, most of
theses tests are based on residual analysis you can easily find such
tests in any book for Econometrics or regression analysis

Best Regards,
Amani