The Role of Explanation Coherence of Two Premises on Property Induction

Three experiments on property induction were conducted to explore whether an incoherent premise discounted the believability of the conclusion when there were two premises. In all three experiments, a single premise increased or decreased the likelihood of the conclusion depending on the nature of explanatory coherence of a premise and a conclusion. However, when there were two premises, one that shared the reason with the conclusion (coherent premise), and one that does not (incoherent premise), the believability of the conclusion was affected differently in three experiments. When two premises and the conclusion were presented simultaneously in Experiment 1, the believability of the conclusion was increased. That is, an incoherent premise did not seem to affect the believability of the conclusion as much as the coherent premise. The incoherent premise seemed to decrease the believability of the conclusion a little bit when the two premises and the conclusion were presented sequentially so that each premise was not ignored in Experiment 2. The incoherent premise decreased the believability of the conclusion below the baseline condition in Experiment 3, where participants were asked to write down reasons for each premise being true. Results of three experiments suggested that only the confirming evidences were processed under natural conditions. A few possible theoretical implications were considered.