This research examined the precision of the pCI rule through three experiments. The results show that first , the tendency of the subjects’ casual judgments was basically similar to the pCI rule. But (a + d) / n predicted human’s casual judgments were even better; second, the increase of subjects’ casual judgments was milder than the pCI rule, and the subjects needed time to construct their own way of judging relationship; finally, different people had different ways of causal judgments, and could be grouped into some categories.