Usual approaches to identify costumer’s perceptions about service are based on subjective evaluations under multiple criteria. Generally, an aggregation function based on weighted average method is used to solve this type of problem, in despite of the compensatory problems that it causes. Present work explores the application of an MCDM based method in order to solve this problem. An original approach to estimates p and q parameters in ELECTRE TRI is inserted on the model building, incorporating robustness to the proposal. Also, an original approach to incomparability’s analyses is observed. An application of the proposal is experimented in a candy shop to evaluate consumer’s satisfaction.
Satisfaction; quality; ELECTRE; multicriteria; evaluation; MDCA
