View article

Linguistic questionnaires are one of the very challenging keys in the world of surveys, in particular considering their fuzziness and imprecision. Many approaches have been used to evaluate them. In this paper, we show the individual and global evaluations of a linguistic questionnaire using fuzzy logic, and the relation between these two evaluations. We explore, as well, the signed distance defuzzification method in the case of different types of fuzzy numbers: the triangular, trapezoidal, gaussian, bell shaped and two-sided gaussian fuzzy numbers. Furthermore, we apply this method to individual evaluations in order to highlight the skewness of the output distributions and compare it to the ones measured using other defuzzification methods. Our simulations revealed some interesting characteristics such as the skewness of distributions obtained from applying the signed distance method is constant for all the types of commonly used fuzzy numbers.