Analysis of the impact of the normalisation technique in the CODAS method on supplier selection

Multi-criteria decision-making methods are applied to various fields of human activity, including the problems of supplier selection. A key step in most of these methods is the choice of the technique used to normalise the decision matrix, i.e. the transformation of input data expressed in different units into dimensionless and comparable numerical values. This paper analyses the effect of the normalisation technique used on the final values and rankings of alternatives (decisions) in the combinative distance-based assessment (CODAS) method in a lumber supplier selection problem. For this purpose, 12 normalisation techniques are used, each consisting of two formulas applied separately to the benefit-type and cost-type criteria. To evaluate these normalisation techniques, a novel use of Hellwig's taxonomic method is proposed, which is then compared with popular measures based on Pearson's correlation coefficient and Spearman's rank correlation coefficient. The results obtained confirm the validity of using the linear normalisation maximum-based method in the combinative distance-based assessment method.