Processing errors in conditional and biconditional problem solving behavior

Abstract

Students ( N = 237) in each of grades 8, 10, 12, and 14 were randomly divided into three groups and administered either a 16-item multiple-choice test of conditional syllogisms, or a 16-item test of biconditional syllogisms, or a 32-item test with both conditional and biconditional syllogisms. Results provided within experiment and within-subjects comparison of responses on conditional and biconditional syllogisms. A comparison of response patterns on conditional items with responses on actual biconditional items provided a direct test of the previously hypothesized biconditional misinterpretation of conditional problems. These analyses confirmed the strong tendency of subjects (across grades) to interpret conditional syllogisms biconditionally. Surprisingly, performance on biconditional problems does not improve systematically with age; in fact, college sophomores perform only slightly better than eighth graders. With regard to conditional syllogisms, results confirmed previously described performance variations across forms of both major and second premises and also replicated an unusual reversed developmental trend on the problem which involves denying the consequent.