Sažetak The main goal of this paper is to explain student's answers on the abstract version of the Wason selection task. Former explanations (Oaksford i Chater, 1994) were based on the assumption that "P" and "Q" cards represent rare events. This assumption is not plausible in all forms of the selection task. Modification of the information gain model is proposed in order to model a students’ performance on the task that involves a conditional proposition (If "P" then "Q"). The model modification relies on the assumption that students construct biconditional representation of the conditional proposition. Advantages of the biconditional interpretation assumption and disadvantages of the rare events assumption are outlined. Experimental manipulations of the conditional proposition task (conditional, standard, biconditional) are included in the tasks with rules where "P" and "Q" elements could not be represented as rare events. Expected changes in a students’ performance on the conditional, standard and biconditional tasks are described. University students (N=142) completed a Wason task in order to put the modified information gain model to empirical test. They were presented with different instructions that suggested conditional, standard and biconditional interpretations of conditional propositions. As the interpretation of the conditional proposition changed, students’ performance on the task also changed in the direction that was predicted by the modified information gain model.
Findings are discussed within a broad framework that includes different versions of the Wason
selection task. The need for to design a general reasoning model that would explain a subjects’ performance on all versions of the selection task is pointed out, and some general directions about the formal properties of such a model are suggested.