Premium
A Logit Model with a Variable Response and Predictors on an Ordinal Scale to Measure Customer Satisfaction
Author(s) -
De Luca Amedeo
Publication year - 2006
Publication title -
quality and reliability engineering international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.913
H-Index - 62
eISSN - 1099-1638
pISSN - 0748-8017
DOI - 10.1002/qre.764
Subject(s) - ordered logit , ordinal regression , ordinal data , logistic regression , econometrics , ordinal scale , statistics , logit , variables , variable (mathematics) , customer satisfaction , coding (social sciences) , scale (ratio) , mathematics , odds , computer science , marketing , business , mathematical analysis , physics , quantum mechanics
A cumulative logit model is presented for assessing the customer satisfaction (CS) of consumers/services users of a product/service. The cumulative logit model (which directly incorporates the order of the categories of the overall CS variable Y response) uses ordinal predictors on an ordinal scale—evaluation judgments on each single attribute of the product—expressed using an additive binary coding scheme provided herein. The scheme naturally identifies the ordering of the categories, does not require the imposition of restraints to avoid non‐monotonous trends of the effect between categories and enables us to achieve more efficient parameters than those obtained by means of the disjunctive binary coding. The model has been assessed using the maximum likelihood method, with the proportional odds hypothesis ( parallel regression ) and with the descending option of the overall dependent variable. This paper suggests the use of the results of the model in an experimental CS planning activity. Copyright © 2006 John Wiley & Sons, Ltd.