A performance indicator and its decomposition according to the impacts of different aspects based on distributional data

In this paper, we present a novel approach to customer satisfaction analysis of airport ser-vices based on the analysis of distributional data for constructing a bivariate performance indicator. Distributional data was introduced for describing macro-data coming from the aggregation of micro-data observed at the individual level. We use them to represent the distribution of the ratings given by 165 classes (macro-units) of airport customers for twelve observed aspects. We describe the trend of passenger satisfaction over time by extracting 165 macro units from a survey conducted among 13,047 passengers at Bari and Brindisi airports during the peak and off-peak seasons of 2015, 2016 and 2017. To obtain a performance indicator, we performed a multiple factor analysis for distributional data. To our knowledge, no other methods exist for the factor analysis of multiple distributional variables. Further, we propose a new visualization tool called Green Eye Iris plot, which allows a joint visualization of our set of distributional values. The obtained results show that the distributional data analysis approach can provide valuable information at macro level that could be hidden when analyzing micro-data or when macro data are represented only by some features coming from summary statistics of groups.