Authors: Odysseas Moschidis, Theodore Chadjipadelis

Title: Testing the Reliability of Satisfaction Measurement Scales using Correspondence Analysis.

Abstract

A k-score evaluation scale, which comprises of hierarchical k degrees, is a measurement scale of the level of satisfaction, behaviour, people’s attitude on various subjects, etc. Testing the level of discrepancy among the results (degrees) of consecutive uses of such a scale in the same sample of people, is an essential precondition for the Reliability of the scale (stability Reliability). Testing the reliability of this scale is also of particular interest as far as the internal consistency of the questions (homogeneity Reliability) used to measure a complex term is concerned. This paper proposes that the stability reliability testing of two successive uses of the scale on a set of questions should be carried out through the application of Correspondence Analysis (CA) on the Comparative Evaluation table . Furthermore, as regards the reliability of the internal consistency (homogeneity) of the questions, we recommend the use of CA on the Evaluation table.