ABSTRACT
This chapter presents three methods for grouping dimensions: Filtered Cluster, which can be used for a small number of interviews; Keyword Network Deconstruction (KND), a more powerful tool based on keyword familiarity that had its origins in a project suggested by Michael Cole during author's 1997–1998 Cattell Fund Fellowship at the University of California San Diego's Laboratory of Comparative Human Cognition (LCHC); and Network Clustering through Ranked and Interpreted Connection Strengths (N-CRIX), an alternative to KND that is less dependent on keyword familiarity. The KND method produced clear groups, but organized the text from most to least common usage. In an educational context, the first groups of dimensions reflected familiar terms, whereas the last groups were based on unfamiliar keywords. The N-CRIX analysis groups half a thousand text samples in a half day of spreadsheet work, but it is complicated enough that only motivated researchers will learn it.
