从Simpson悖论看群体生物学中统计解释和因果解释之争

在理论的群体生物学中,有三种模型都是通过性状适合度(trait fitness)的分布来解释和预见种群结构变化的量和方向。双因子和单因子模型都将性状适合度看成是一种因果属性,而丹尼斯.沃尔什(Denis Walsh)等则指出它们仅仅是一种统计学的属性。笔者通过分析沃尔什的有力论证,尝试表明如果将性状适合度看成是一种因果属性的话,将导致Simpson悖论,因而是不融贯的。所以,关于种群结构变化的统计学解释要优于因果解释。

Studies on the Controversy between the Statistical Explanation and Causal Explanations in Population Biology from the View of Simpson Paradox

Within the population biology,there are three models use trait fitness to explain and predict the change of the structure of a population.Both the two-force model and single-force model regard the trait fitness as a kind of causal properties.However,Denis Walsh takes it as statistical.I try to show the causal explanation of the trait fitness will face up to Simpson paradox and therefore it is inconsistence by spelling out the Denis’s convincing arguments which argue against the causal explanation of trait fitness.Thus,the statistical explanation has some advantages over the causal explanation of the trait fitness.