Ordering of information in conditional reasoning

Two experiments are reported in which participants are asked to evaluate computer-presented conditional syllogisms consisting of a major premise (conditional rule), a minor premise and a conclusion. As an example the modus tollens inference with an affirmative conditional has the form if p then q, not-q therefore not-p. In these experiments, participants were asked to judge the validity of four types of conditional syllogisms with four types of conditional rule in which the presence and absence of negated components was varied. In Expt 1, half of the participants received problems with the conclusion presented prior to the premises, and half in the usual order. Experiment 2 combined this variable with a second: standard or reversed order of major and minor premises. Both manipulations substantially reduced the effects of the negative conclusion bias which is typically observed in this paradigm. The findings are discussed with respect to the debate between mental models and mental logic accounts of conditional inference.