Computationally intelligent methods for qualitative data analysis

Abstract

This study focuses on computationally intelligent methods, which are applied to the analysis of survey data in educational research. The methods can be used with complex data sets, which contain several data types. Each data type is analyzed in a separate subanalysis, and the results from these subanalyses can be combined. The methodology makes it possible to locate groups of similar answers from the subanalyses, and to identify these groups using background information. It also allows one to compare groups that are selected from different subanalyses, from different populations, and to locate and identify similar textual answers. In connection to this study, a software application has been created to test the developed methods.