Signed difference analysis: A brief introduction

The following is a introduction to a method, called signed difference analysis (or SDA), that I have been developing, off and on, over the last seven years. SDA is a general technique for evaluating psychological models using only the “non-metric” or ordinal properties of dependent variables. Although still a work in progress, the following is an outline of its main elements which, hopefully, will communicate the basic idea of SDA and give some indication of how and where it can be applied. The article is divided into four parts. In the first part, I discuss the central problem of psychological measurement that SDA attempts to address. In the second part, I present a concrete example of a psychological model and, in the third part, show how SDA can be used to derive predictions that use only ordinal information. In the final part, I discuss the possible application of SDA to cognitive neuropsychology.