DAM Weighting: A New Approach to Determining Criteria Weighting in Multi-Criteria Decision Making

The weighting of criteria in multi-criteria decision-making (MCDM) is essential because it reflects the level of importance or priority of each criterion considered in decision-making. This paper aims to propose an innovative objective weighting method called the Data Assessment Model (DAM) to increase the validity and reliability of objective weighting in multi-criteria decision-making. The DAM method enables the weighting of criteria to be determined based on empirical data, rather than solely on a subjective assessment made by one or several decision-makers. The Pearson correlation analysis was conducted to measure the linear relationship between the ranking vectors of alternatives generated by different combinations of objective weighting methods and MCDM techniques, including SAW, MOORA, TOPSIS, and SMART. In this context, each correlation value represents the similarity between alternative rankings produced using different weighting schemes within the same MCDM method. The results show that, in combination with SAW, the MEREC and DAM methods achieved a perfect correlation value of 1, followed by CRITIC with 0.909 and Entropy with 0.825. For TOPSIS, DAM demonstrated the highest correlation value of 0.902, followed by MEREC with 0.881, while Entropy and CRITIC showed lower values of 0.811 and 0.825. In the MOORA method, DAM again reached the maximum value of 1, followed by MEREC with 0.993 and CRITIC with 0.95, whereas Entropy remained lower at 0.825. Meanwhile, for SMART, CRITIC produced the highest correlation value of 0.979, followed by DAM with 0.965, MEREC with 0.95, and Entropy with 0.888. These results indicate that the DAM and MEREC weighting methods tend to produce ranking patterns that are more consistent with other methods, while the Entropy method shows relatively greater variation. From these results, it can be concluded that the DAM Weighting method has the strongest correlation with the tested data, while the Entropy method shows a relatively lower correlation compared to the other methods.