The article uses data from two experiments on the content and accuracy of mental representations of space by the blind and visually impaired in order expose some of the shortcomings of typical statistical testing and propose an individual differences approach to the analysis of data.
It begins with a discussion of some of the problems associated with the strict classification and eventual comparison of individuals between groups. The individual differences approach is then presented and the concepts of ability and present competence are explored along with the importance of detailed participant description. Examples from the two experiments are used to demonstrate how null hypothesis significance testing can be complemented with effect size estimates, box-plots and ranking techniques. Throughout the article we are reminded of the need to adopt mutually supportive techniques to account for the heterogeneity of experience and skills between participants.
This working paper is available as a PDF. The file size is 948KB.
Authors: Victor Schinazi
Publication Date: 1/7/2006