Fuzzy Likert scales

Aggeliki Kazani
Description

The present paper studies the fuzzification of Likert scales. Closed type questions-units which are used for the construction of these particular scales oblige the respondent to choose only between the available options that may not reciprocate to the real opinion or attitude. The theory of fuzzy logic and fuzzy systems is the tool that will be used for the fuzzification of Likert scales. This paper’s aim is the fuzzification of the Likert scale “Macedonia and the Balkans, Xenophobia and development”, published by the National Centre for Social Research which was included in the survey to explore opinions and attitudes on social stereotypes about the “foreigns-others”. In order to fuzzificate the Likert scale three methological steps were used. Initially, after the development of operational definitions and discovering the relation between the questions-units, a hypothetical fuzzy system was constructed. Then, fuzzification of input variables was realized by examining the bar charts of each variable individually and presenting them by fuzzy partitions, giving priority to finding the structure and the relationship of those variables. Thus, using the empirical knowledge of the experts, rule-blocks were constructed that led to the description of system of xenophobia, which proves the vagueness of this particular Likert scale and provides an alternative approach of the measurement of xenophobia.

Supervisor
Thesis Committee Members
  • Prof. A. Kalamatianou
Academic Year
2009 - 2010