Methodological limitations of linear parametric analysis in biological research: A critical review of NEO-Five personality traits and sleep characteristics study

This paper critically examines the methodological approach employed in Wang et al.'s systematic review and meta-analysis of NEO-Five Personality Traits and sleep characteristics. We identify significant concerns regarding the use of linear parametric methods (Pearson's correlation and meta-regression) in analyzing potentially nonlinear and nonparametric biological data. The paper demonstrates how these analytical tools can lead to distorted results and erroneous conclusions when applied to complex biological systems. We propose alternative nonlinear nonparametric methods, including Spearman's correlation and Kendall's tau for monotonic relationships, and Mutual Information analysis and Effective Transfer Entropy for nonmonotonic interactions, to better capture the intricate relationships in biological research without restrictive assumptions about linearity or data distribution.