A Framework for Recognizing Mechanistic Reasoning in Student Scientific Inquiry

Thumbnail Image


umi-umd-3941.pdf (7.14 MB)
No. of downloads: 1409

Publication or External Link






A central ambition of science education reform is to help students develop abilities for scientific inquiry. Education research is thus rightly focused on defining what constitutes "inquiry" and developing tools for assessing it. There has been progress with respect to particular aspects of inquiry, namely student abilities for controlled experimentation and scientific argumentation. However, we suggest that in addition to these frameworks for assessing the structure of inquiry we need frameworks for analyzing the substance of that inquiry.

In this work we draw attention to and evaluate the substance of student mechanistic reasoning. Both within the history and philosophy of science and within science education research, scientific inquiry is characterized in part as understanding the causal mechanisms that underlie natural phenomena. The challenge for science education, however, is that there has not been the same progress with respect to making explicit what constitutes mechanistic reasoning as there has been in making explicit other aspects of inquiry.

This dissertation attempts to address this challenge. We adapt an account of mechanism in professional research science to develop a framework for reliably recognizing mechanistic reasoning in student discourse. The coding scheme articulates seven specific aspects of mechanistic reasoning and can be used to systematically analyze narrative data for patterns in student thinking. It provides a tool for detecting quality reasoning that may be overlooked by more traditional assessments.

We apply the mechanism coding scheme to video and written data from a range of student inquiries, from large group discussions among first grade students to the individual problem solving of graduate students. While the primary result of this work is the coding scheme itself and the finding that it provides a reliable means of analyzing transcript data for evidence of mechanistic thinking, the rich descriptions we develop in each case study help us recognize continuity between graduate level learning and elementary school science: part of what students are able to do in elementary school finds its way to graduate school. Thus this work makes it possible for researchers, curriculum developers, and teachers to systematically pursue mechanistic reasoning as an objective for inquiry.