Novel approach in data aggregation and its application in airline service evaluation

Abstract

The use of questionnaire data has been a common method to gain more insights from service or product users, however interpreting this kind of data is never a straight-forward task and sometimes it can lead to ambiguous or misleading results. This thesis paper suggests a new approach for the evaluation and analysis of questionnaire data. It suggests a multiple criteria multi-expert evaluation framework where importance of specific answers from respondents will be reflected. Likert-scale inputs and aggregated evaluations are represented using histograms. The concept of histogram weights of criteria for the analysis of questionnaire data is introduced. The proposed method was empirically tested on the real-world data set, evaluation of Vietnamese airlines by their customers in 2017. The result shows that we can obtain a lossless representation of the evaluations of a given alternative by a group of evaluators overall , get the marketing implications of the lossless representation of questionnaires and reflect weights of answers specified by the user/buyer of the analysis.. The concept of a 2-dimensional histogram to represent the data with subjective importance of the partial evaluations as expressed by the experts and propose histogram weights is a useful tool for the analysis of respondents’ preferences in marketing research.