


A model of data is a mathematical description of the relationships between two or more variables  in essence, a statistical model (Dawson, 2004; Lunneborg, 1994). For example, a multiple regression equation that predicts the value of an independent variable from the weighted values of one or more dependent variables is a model of data. In general, statistical methods represent models of data. These models are used to provide compact descriptions of data which can then be communicated to others. Usually they are linear, and are evaluated in terms of their goodness of fit to a set of observations that have been modeled. They are important because some of their properties (linearity, evaluation via goodness of fit), do not apply to other types of models (such as connectionist networks), which indicates that researchers have a wide variety of different types of models from which to choose to suit their particular modeling needs.
References:
 Dawson, M. R. W. (2004). Minds And Machines : Connectionism And Psychological Modeling. Malden, MA: Blackwell Pub.
 Lunneborg, C. E. (1994). Modeling Experimental And Observational Data. Belmont, CA: Duxbury Press..
(Added October 2009)



