Normalization methods for spatio‐temporal analysis of environmental performance: Revisiting the Min–Max method
Matteo Mazziotta,Adriano Pareto
DOI: https://doi.org/10.1002/env.2730
2022-05-24
Environmetrics
Abstract:Over the last few years, composite indices for ranking or assessing country performance in a wide variety of complex phenomena, such as environmental sustainability, have gained a lot of popularity. However, not all indices allow for detailed spatio‐temporal analyses. For instance, if individual indicators to be aggregated are normalized by the Min–Max method, they are converted to a common scale with a range of [0, 1] and it can be difficult to appreciate any absolute change in country performance at the extremes of the range (i.e., when a country scores 0 to 1). In this article, an alternative method for normalizing data in a three‐way array of the type units × variables × times, is considered. It normalizes the range of individual indicators, similarly to the Min–Max method, but uses a common reference that allows to "center" them, without forcing them into a closed range. An application to renewable energy consumption data is also shown.
environmental sciences,statistics & probability,mathematics, interdisciplinary applications