The near totality of human mate choice research occurs in large-scale, urban, industrial populations. It is unclear to what extent lessons learned from such populations reflect generalizable features of human mating psychology as opposed to localized responses to the demands of these historically unusual environments. Here, we use couple simulation, an agent-based modeling technique, to compare models of mate choice across both a U.S. sample (n = 1678) and a sample of k = 15 couples from Conambo, Ecuador—a relatively remote community of horticultural-foragers in the Ecuadorian Amazon. The Conambo sample provides a unique opportunity to evaluate models of mate choice in that (1) this sample represents approximately 50% of the households within this community and (2) all of the participants in this sample are acquainted with one another. Participants in Conambo completed a ranking task in which each participant ranked each opposite-sex adult in the community in terms of their quality as a spouse. We used these rankings to simulate the mating market in Conambo under alternative models of mate choice. We find that these models are able to reproduce Conambo marriages at a high degree of accuracy and perform comparably across both the Conambo sample and U.S. samples. Specifically, the resource allocation model performs best in reproducing mate choices in both the U.S. and Conambo samples. These results suggest that at least some aspects of human mating psychology generalize across both large-scale industrialized and small-scale populations.

人类择偶研究几乎全部发生在大规模的城市工业人口中。目前尚不清楚从这些人群中学到的经验教训在多大程度上反映了人类交配心理的普遍特征，而不是对这些历史上不寻常环境的要求的局部反应。在这里，我们使用情侣模拟（一种基于主体的建模技术）来比较美国样本 ( n  = 1678) 和 来自厄瓜多尔科纳博（一个相对偏远的园艺社区）的k = 15 对情侣的择偶模型。厄瓜多尔亚马逊地区的采集者。Conambo 样本提供了一个评估择偶模型的独特机会，因为 (1) 该样本代表了该社区内大约 50% 的家庭，并且 (2) 该样本中的所有参与者都彼此熟悉。Conambo 的参与者完成了一项排名任务，其中每个参与者根据配偶的质量对社区中的每个异性成年人进行排名。我们使用这些排名来模拟科纳博在择偶选择模型下的交配市场。我们发现这些模型能够高度准确地再现 Conambo 婚姻，并且在 Conambo 样本和美国样本中的表现相当。具体来说，资源分配模型在再现美国和科南博样本中的配偶选择方面表现最佳。这些结果表明，人类交配心理的某些方面至少适用于大规模工业化人口和小规模人口。