Fonte: Journal of Tourism & Development; v. 1, n. 17/18 (2012); 87-96
Palavras-chave:
Resumo: Ordinal logistic regression models have come to be used in the analysis of data whose response variable is of ordinal type meaning that the categories of the variable can be ranked but the distances between categories are unknown. In studies of customers’ satisfaction it is usual to measure the satisfaction of the customers or the customers´ recommendation of the service by means of ordinal scales. This paper presents a description of the proportional odds model and the partial proportional odds model. Although the existence of quite a lot studies on customers´ satisfaction and loyalty of customers in the hotel industry, the use of this type of models in this area is still reduced. The theoretical framework of the main ordinal regression models is presented, as well as their utilization conditions. The use of this sort of models is illustrated with an empirical study using a set of customer’s answers of a hotel unit in the Algarve region. It was selected the ordinal regression model that reveal adequate to model the data. As main conclusions to relate is that the age of the customers and the duration of the stay have influence in customers´ intention to recommend the service.
