Management questionnaire analysis through a linguistic hard C-means

Management studies often gather abstract, numerical data from humans. Employees' attitudes and perceptions regarding organizations and their practices are collected. Most models that use this data account for probabilistic uncertainty, but not for vagueness uncertainty. In an earlier paper, Adrian (1998) considered the utility of allowing respondents to draw fuzzy membership functions over the set of questionnaire answers instead of just picking one response. Her results demonstrated that good qualitative information could be obtained from this format. In this paper, we consider a quantitative analysis of these linguistic responses. In particular we develop a computationally efficient linguistic hard C-means algorithm for vectors of fuzzy sets and apply this algorithm to such linguistic vectors derived from a set of subjects answering questions about job satisfaction and organizational commitment.