Neutrosophic MULTIMOORA: A Solution for the standard error in information sampling

Abstract

If complete Data Mining is not possible one has to be satisfied with an information sample, as much representative as possible. The Belgian company “CIM” is doing marketing research for all Belgian newspapers, magazines and cinema. For some local newspapers, it arrives at a standard error of more than 15% or a spread of more than 30%, which is scientific nonsense but accepted by the publishers of advertisement. On the other side technical problems will ask for a much smaller standard deviation like for instance a standard error of 0.1% for the possibility that a dike is not strong enough for an eventual spring tide. Somewhat in between the usual standard error for marketing research is 5%. Is it possible to avoid this Spread by Sampling? Here Multi-Objective Optimization Methods may help. The Neutrosophic MULTIMOORA method, chosen for its robustness compared to many other competing methods, will solve the problems of normalization and of importance, whereas Fuzzy MULTIMOORA may take care of the annoying spread in the marketing samples. While an application on the construction of dwellings is given, many other applications remain possible like for Gallup polls concerning public opinion, general elections in particular.