“…Although Kendall τ and Spearman ρ offer some advantages by relaxing the normal distribution assumption, they do not account for the possible presence of temporal trends, which can strongly affect the correlation value and yield artefactual associations. More sophisticated methods such as spatial ( Andersen, et al, 2021 ; Maroko, Nash, & Pavilonis, 2020 ; Parvin, et al, 2021 ) or generalized linear models for count data, including Logistic ( Kwok, et al, 2021 ; Birhanu, Ayana, Bayu, Mohammed, & Dessie, 2021 ), Poisson ( Sugg, et al, 2021 ; Morrissey, Spooner, Salter, & Shaddick, 2021 ) and Negative Binomial ( Benita & Gasca-Sanchez, 2021 ; Strully, Yang, & Liu, 2021 ), have been also applied. Machine learning models and classification algorithms figure from amongst the most popular approaches for predicting COVID-19 occurrence using socioeconomic inputs ( Phiri, et al, 2021 ; Li, et al, 2021 ).…”