Do you know that you can create a vector that has a specific correlation with another vector? That is, given a vector, x, and a correlation coefficient, ρ, you can find a vector, y, such that corr(x, y) = ρ. The vectors x and y can have an arbitrary number
To help visualize regression models, SAS provides the EFFECTPLOT statement in several regression procedures and in PROC PLM, which is a general-purpose procedure for post-fitting analysis of linear models. When scoring and visualizing a model, it is important to use reasonable combinations of the explanatory variables for the visualization. When
Intuitively, the skewness of a unimodal distribution indicates whether a distribution is symmetric or not. If the right tail has more mass than the left tail, the distribution is "right skewed." If the left tail has more mass, the distribution is "left skewed." Thus, estimating skewness requires some estimates about
