SCIENTIFIC THINKING ABOUT SCIENTIFIC THINKING

Readers come to the Commentaries of these Monographs with three questions in mind. First, is the Monograph worth reading? Second, does it raise specific points that deserve further attention, emphasis, or criticism? Finally, are there broad issues raised by the Monograph that are sufficiently important to warrant further discussion, almost independent of the content of the Monograph itself? With respect to the present Monograph, the short answers to these questions are "yes," "yes," and "yes." The longer answers follow. address the question of how people generate evidence about multivariable causal systems and then form hypotheses about the relevant variables on the basis of that evidence. They investigated this issue in two broad domains (physical and social), using two groups of subjects (preadolescents and adults), and they focused on how, over the course of 10 weekly experimental sessions, subjects acquired not only domain-specific knowledge (e.g., the factors that make for fast cars or effective television commercials) but also domain-general strategies for making valid inferences from data. The design enabled Kuhn et al. to compare performance within and across domains and subject populations, and the analysis revealed important similarities and differences in the use of valid and invalid strategies that we will describe below. Kuhn et al.'s ambitious and unprecedented undertaking embraces a densely interwoven tapestry of fundamental methodological issues and central topics within the area of cognitive development. The methodological issues include transfer of training, microgenetic analysis, and the relative merits of quantitative and qualitative analysis of children's behavior. The 137