The vast majority of applications of statistics in the social sciences have a tendency to ignore two simple facts:

1) The techniques assume the underlying data are normally distributed and

2) Most data are not normally distributed.

Further, we often assume that our search for relationships among data must be constrained to linear ones. For example, the most common and often misused term related to data analysis is correlation, but what does it really mean? This session briefly discusses why we need to get away from normally distributed data and linearity assumptions and offers one possible solution to this problem. Data from a recent validation study is presented to show how successful this proposed solution has the potential to be.