Rethinking data presentation in trace metal studies of food and environmental matrices: Consequences for statistical inference, reproducibility, and risk assessment

Abstract

Trace metal analysis in food and environmental matrices is central to environmental monitoring, food safety evaluation, and public health risk assessment. However, the value of these data is often limited by inconsistent statistical treatment and reporting practices. Trace metal datasets are inherently heterogeneous, frequently censored near analytical detection limits, and commonly right-skewed, leading to routine violations of classical statistical assumptions. When these characteristics are not transparently addressed, interpretation, comparability, and evidence synthesis across studies are compromised. This paper presents a synthesis-driven, perspective-based methodological review of statistical and reporting practices in trace metal studies of food and environmental matrices. The review critically evaluates why common analytical approaches are used, how alternative methods compare, and where their limitations affect inference and risk interpretation. Emphasis is placed on data transformation and back-transformation, interpretation of negative log-transformed values, treatment of censored data, and reporting of concentration metrics, including detection limits and wet-weight versus dry-weight bases. Log transformation is widely applied to address skewness and enable parametric analyses, yet its rationale and interpretive consequences are often insufficiently explained. Negative log-transformed values are frequently misinterpreted, reflecting communication gaps rather than analytical errors. In addition, inconsistent or undocumented reporting of concentration weight basis substantially limits cross-study comparability, reproducibility, exposure assessment, and meta-analysis. This perspective argues that data presentation in trace metal research is a methodological and ethical responsibility shaping inference, comparability, and decision-making. Practical best practices are proposed, including diagnostic-driven transformation decisions, dual presentation of transformed and original-scale data, standardized reporting of detection limits and weight basis, explicit justification of statistical methods, and routine data sharing. Adoption of these practices will strengthen reproducibility and enhance the contribution of trace metal research.