A New Framework for Teaching Scientific Reasoning

Abstract

I have been teaching a lower division course on scientific reasoning for roughly thirty years. After the first ten years I wrote a textbook, Understanding Scientific Reasoning (1979), based on the course as it had developed over the previous decade. The overall framework of this text followed a pattern going back to Mill’s Logic (1843) and perpetuated up to the present. It began with deductive logic and then took up scientific reasoning understood as a kind of inductive logic. What mainly differentiated my text from others was that less than one fourth of the book was devoted to deductive logic and more than one fourth focused on statistical reasoning as understood by a majority of contemporary statisticians. I deliberately did not introduce Mill’s Methods. Nevertheless, it shared with the vast majority of texts in logic and reasoning, whether formal or informal, the assumption that the evaluation of any particular bit of reasoning is done by first reconstructing that reasoning as an explicit argument, with premises and a conclusion, and then examining the reconstructed argument to see if it exhibits the characteristic form of a good argument, whether deductive or inductive. In the third (1991) and fourth (1997) editions, I have abandoned this assumption. From the beginning, Understanding Scientific Reasoning was designed to impart the intellectual skills of understanding and evaluating scientific reasoning as presented in print media ranging from semi-professional technical journals to tabloid newspapers. It was not intended to be a textbook in the philosophy of science. It contains, therefore, almost no discussions reflecting on the framework employed to teach the desired skills. The framework is simply presented and employed. This paper, intended for teachers rather than students of scientific reasoning, attempts to explain, and to some extent justify, the new framework.