Categorization Methodology: an Approach to the Collection and Analysis of Certain Classes of Qualitative Information

A new methodology has been developed for the study of certain classes of qualitative information. The methodology is composed of two major techniques: the F-sort task for data collection and latent partition analysis for data summarization. In this paper, similarities and differences of the methodology in relation to existing techniques are discussed. This includes a review of the historical antecedents and development of the methodology as well as a listing of recent applications of the methodology, which have been in several fields of psychology and education. Following that review, a detailed presentation is given of how the methodology was applied to studying teachers' views of facilitating student learning in the classroom. In the data analysis, some previously unpublished enhancements of latent partition analysis are introduced -- for basic categorization statistics, recovery of subgroup information, and linkage to multidimensional scaling. This complete and substantively intriguing example is intended to illustrate the power and utility of the methodology in exploring important research topics. Then, examinations are made of the main components of the F-sort technique and of latent partition analysis. The intent is to lay out detailed frameworks for designing and interpreting research that applies the categorization methodology. The methodology has not previously been given an integrated description. Finally, discussion concerns the uses of the methodology and the importance of standardized application of the procedures.