Form and content in the development of deductive reasoning.

Abstract

Three experiments explored the development of formal logical reasoning between Grades 4 and 12 and the role of semantic content in the solution of Wason's (1966) selection task problems. In Experiment 1, subjects in the 8th, 10th, and 12th grades were tested on several familiar-content and several abstract-content selection problems. In Experiment 2, subjects in Grades 4, 8, and 12 were assessed on the familiar problems used in Experiment 1. In Experiment 3, subjects in Grades 4,6, and 8 were tested on novel familiar-content problems, familiar-content problems with conditional clauses reversed, a meaningful but unfamiliar problem, and an abstract problem. The results suggested that formal logical reasoning is not generally present during the 4th or 6th grades and that formal logical competence becomes available in adolescence. A general facilitation effect was found for familiar semantic content, but this was qualified by one anomalous familiar-content problem in each experiment. The results are discussed in the context of the competence-moderator-performance model, which maintains that both logical knowledge and world knowledge are necessary but distinct features of adequate reasoning performance. A major issue in the field of deductive reasoning concerns the role of logic versus the role of semantic content in determining performance on reasoning problems. One theoretical approach (Overton, 1985;Staudenmayer, 1975) maintains that the development of a logical competence is a necessary prerequisite for solutions to deductive reasoning problems. According to this approach, specific content effects operate as moderators of logical competence to inhibit or enhance successful task performance. From this perspective, both formal reasoning processes and the real-time information-processing strategies related to content effects play essential roles in determining task performance. A second theoretical approach (Griggs, 1983; Mandler, 1983) asserts that formal reasoning processes play little or no role in task performance. According to this approach, task performance is virtually totally determined by context variables such as those provided by specific semantic contents. The aim of the present research is to explore both the role of the development of formal logical reasoning and the role of semantic content in the solution of deductive reasoning problems. The development of formal logical reasoning is represented in Piaget's theory (Inhelder & Piaget, 1958) as a succession of increasingly powerful logical systems or competencies. Several recent studies (Byrnes & Overton, 1986; O'Brien & Ov