样例学习条件下的因果力估计

CAUSAL JUDGMENTS IN THE TRIAL-BY-TRIAL PRESENTATION

在逐个呈现因果样例的条件下 ,考察单一因果关系因果力估计的特点 ,同时检验联想解释 ,概率对比模型 ,权重ΔP模型 ,效力PC理论和pCI规则。实验让 6 5名大学生被试估计不同化学药物影响动物基因变异的能力。实验结果表明 :(1)对产生原因的因果力估计符合权重ΔP模型 ;(2 )对预防原因的因果力估计较多符合效力PC理论 ;(3)因果力估计具有复杂多样性 。

An experiment was conducted to investigate human causal judgments in the trial-by-trial presentation, and to test the associative theory, the power PC theory, the probabilistic contrast model, the weighted ΔP model, and the pCI rule. The participants were 65 college students. It was found that causal judgments had three characteristics: (a) Causal judgments were in accord with the weighted ΔP model in generative causes, (b) more causal judgments were in accord with the power PC theory in preventive causes, and (c) causal judgments were difficult to be described and generalized with a uniform pattern.