Comparison of Fuzzy and Statistical Approaches in Multicriteria Decisionmaking

Abstract Criteria used to measure ecological or social sustainability of forestry often include qualitative ones. Fuzzy additive weighting has been proposed for using qualitative data in the form of linguistic variables. It is, however, possible to formulate a linguistic problem also using statistical distributions. We compared the fuzzy and statistical approach in a linear additive weighting problem on a woodlot owned by UPM-Kymmene. The results of fuzzy additive weighting showed high uncertainty concerning the decision compared to the statistical approach. In fuzzy analysis, the best option was unclear and the risk involved in the decision seemed high, while in statistical analysis the best option was clear and the risks involved seemed negligible. The uncertainty present in the fuzzy approach did not correspond to the 95% confidence intervals, but rather to the extreme values of statistical analysis. It can be concluded that the fuzzy approach represents ignorance in the true criterion values, while the statistical approach represents variability in them. It can also be concluded that the fuzzy approach represents decisionmaking under uncertainty, while the statistical approach represents decisionmaking under risk. If probability distributions for the uncertain data can be obtained, the statistical approach is more useful. Otherwise, the fuzzy approach should be used.