Abstract: Property effect, also called interaction between properties and premise-conclusion matches, is an important psychological effect in inductive reasoning, which suggested that people could base on the background relation (relations between premise and conclusion) property most relevant to. At the same time, previous studies showed that different types of background relation vary in their relative accessibility. Therefore, the current study tries to examine which kind of background relation people would give priority to, the most accessible or the property most relevant, in inductive reasoning.
The experiment consisted of two parts: a property induction task followed by a belief- assessment task, both were conducted on the computer using E-Prime. In the induction task, 64 Chinese undergraduates were told about a novel gene or a novel disease that was true of one category of animals, they had to judge whether taxonomically, ecologically, and unrelated animals had the same property or not, under speeded or delayed conditions. Under speeded conditions the second category of animals presented 1s (delayed conditions 15s), participants were asked to judge as quickly as possible after they disappeared. After the judgment, they were asked to evaluate the confidence in their judgment. In the belief-assessment task, participants were shown each item again and asked “Do these animals live in the same habitat?” and “Do these animals belong to the same biological category?” They answered “yes” “no” or “don’t know”.
The result revealed that: (1) Property effect was independent of time pressure. The possibility of reasoning(the percent of judged “yes” in the induction task) for gene between taxonomically related animals was greater than for disease in both speeded and delayed conditions, but between ecologically related animals, there was no difference. (2) Time pressure affected the confidence of reasoning, higher in speeded conditions than in delayed conditions. (3) In the speeded conditions, both taxonomical and ecological relatedness beliefs could predict the possibility and confidence of reasoning, but in delayed conditions both beliefs couldn’t.
These results proved that people are firstly extract the background relation property most relevant to then extract those less relevant or un-relevant in inductive reasoning.

Key words: inductive reasoning, property effect, time pressure