ПРЕПРИНТ

Эта статья является препринтом и не была отрецензирована.
О результатах, изложенных в препринтах, не следует сообщать в СМИ как о проверенной информации.
Practical application of the Wilcoxon-Mann-Whitney test in valuation
2025-04-07

Appraisers often face the need to take into account differences in quantitative and qualitative characteristics of objects in their practice. One of the standard tasks is to determine the attributes that influence the cost (so-called “pricing factors”) and to separate them from the attributes which influence do not or cannot be determined, in particular. Subjective selection of attributes taken into account in determining the value is widespread in valuation practice. In this case, specific quantitative indicators of the impact of these attributes on the cost are regularly taken from the so-called “reference books”. While not denying the speed and low cost of this approach, it should be recognized that only data directly observed in the open markets is a reliable basis for a value judgment. The priority of such data over other data, in particular those obtained by expert survey, is enshrined, among others, in RICS Valuation — Global Standards 2022, International Valuation Standards 2022 as well as in IFRS 13 “Fair Value Measurement”. Therefore, we can say that mathematical methods for analyzing data from the open market are the most reliable means of interpreting market information used in market research and predicting the value of individual objects. The aim of this paper is to justify the necessity and possibility of using a rigorous mathematical Wilcoxon– Mann– Whitney test. This test allows us to answer the question about the necessity of taking into account the binary attribute as a price-generating factor. At the time, the judgmental approach is most commonly used by appraisers in selecting the attributes to be considered in appraisal. But this paper proposes the idea of prioritizing the measuring approach based on the results of a mathematical test. It allows drawing a conclusion about the importance or otherwise of the binary attribute influence on the value. There is an important fact to note. The test under consideration is based on a frequentist approach to probability. However, it is also relevant to machine learning methods through its connection to ROC analysis and the concept of AUC. The issues of such a relationship will be discussed in this paper. The existence of such a relationship seems very interesting in terms of introducing elements of the Bayesian approach to probability into the practice of appraisers. Users should have some general math background and basic Python and R programming skills to understand and practice all the material in the text. But some lack of that knowledge and skill is not a barrier to learning most of the material and implementing the test in the spreadsheet. The material consists of four blocks: — a general description of the Wilcoxon– Mann– Whitney test (hereafter “the U-test”), its probabilistic meaning and its relationship to other mathematical methods; — a practical implementation of the U-test in a spreadsheet on an example of test random data; — practical implementation of the U-test on the real data of the residential real estate market of St. Petersburg agglomeration by tools of Python programming language; the purpose of the analysis was to check the significance of the difference in the unit price between the objects located in the urban and suburban parts of the agglomeration; — practical implementation of the U-test on real data of residential real estate market of Almaty by tools of R programming language; the purpose of the analysis was to check the significance of difference in unit price between the objects sold without demountable improvements and chattels and the objects sold with them. The current version of this material, its source code, Python and R scripts and the spreadsheet are in the repository on the GitHub portal and available at permanent link. This material and all of its appendices are distributed under the terms of the cc-bysa-4.0 license.

Ссылка для цитирования:

Murashev C. A. 2025. Practical application of the Wilcoxon-Mann-Whitney test in valuation. PREPRINTS.RU. https://doi.org/10.24108/preprints-3113493

Список литературы