Quantified reasoning in formal specification : transfer of everyday errors and biases?

Abstract

Cognitive studies of syllogistic reasoning appear to provide important pointers to the ways in which people reason with quantified statements. Previous natural language based studies of the syllogistic task suggest that novice reasoners are prone to systematic errors and biases. In this report, we discuss some of the cognitive explanations for these and report a study aimed at testing whether computing scientists with the relevant training in logical deduction and the Z formal notation are liable to succumb to the same non-logical tendencies when reasoning about categorical syllogisms expressed in Z. The results suggest that many of the errors and biases which people exhibit on a frequent basis when reasoning about quantified statements in everyday communication can transfer over into the formal domain. The implications of this finding are discussed in relation to the software engineering community where formal specifications are becoming increasingly used in the development of business and safety-critical systems.