Watson Elijah J, Glass Delaney J, Petito Lucia C
Department of Anthropology, Northwestern University, Evanston, Illinois, USA.
Department of Anthropology, University of Toronto - St. George, Toronto, Ontario, Canada.
Am J Hum Biol. 2025 Sep;37(9):e70149. doi: 10.1002/ajhb.70149.
Human biologists seek to understand how cultural, environmental, and biological forces shape observed patterns of human variation. Yet contemporary insights and approaches to observational causal inference remain underutilized in the field. We outline a structured but flexible roadmap for causal inference in human biology that begins with theory development, defines causal questions and estimands, employs directed acyclic graphs (DAGs) to clarify assumptions, and evaluates key identification criteria prior to statistical analysis. We position this framework within a spectrum of causal inference traditions, spanning from interventionist approaches rooted in well-defined, manipulable exposures to realized approaches that engage historically situated and ecologically embedded phenomena. Rather than offering a prescriptive checklist, we frame this toolkit as an opening: a step toward anthropological causal inference that integrates transparency, theoretical and methodological coherence, and the epistemological commitments of the biocultural synthesis in human biology and anthropology.
人类生物学家试图了解文化、环境和生物力量如何塑造所观察到的人类变异模式。然而,当代关于观察性因果推断的见解和方法在该领域仍未得到充分利用。我们概述了一个用于人类生物学因果推断的结构化但灵活的路线图,该路线图始于理论发展,定义因果问题和估计量,使用有向无环图(DAG)来阐明假设,并在统计分析之前评估关键识别标准。我们将这个框架置于一系列因果推断传统之中,从植根于明确、可操纵暴露的干预主义方法到涉及历史情境和生态嵌入现象的现实方法。我们并非提供一份规定性的清单,而是将这个工具包视为一个开端:迈向整合透明度、理论和方法连贯性以及人类生物学和人类学中生物文化综合的认识论承诺的人类学因果推断的一步。