Information about ongoing research at the lab.

Research by Chris Westbury and Geoff Hollis

The Naturalistic University of Alberta Nonlinear Correlation Explorer  (NUANCE) is a tool that allows you to harness the power of evolution in  your attempts to explore intricacies within your data. The conventions  of psychology typically dictate that we treat variables categorically  during analysis (e.g. high/low event frequency, regular/irregular word  spelling). However, many of the factors we use to study psychological  phenomena are continuous, and our treatment of them as categorical  indicators can gloss over many of the finer grained details of what is  going on. NUANCE allows you to build mathematical representations of  your variable relations, defined across each variable's entire range.  screenshotHowever, unlike most regression techniques in common use, NUANCE's  ability to provide suitable fits does not require you to have prior  knowledge of either the proper transformations that need be applied to  the data, or of the shape of the curve relating your predictor to your  dependent variables. NUANCE is the computational embodiment of  selective breeding: with a pressure towards minimizing error, NUANCE  manages thousands of mathematical formulae that describe your data,  selecting those that fit it best and recombining them in random ways to  generate new formulae. Through this iterative process, NUANCE will show  you ways to look at your data you would have never imagined, and  highlights important relationships you would have never assumed existed!

For more information about NUANCE, click here.
Download NUANCE (free for academic use).
Read the NUANCE documentation before downloading.


©2005  WestburyLab   chrisw at ualberta dot ca