Learning Scientific Principles With Contrasting Cases: Key Ingredients of Effective Contrast-Focused Instruction

• In Event: On the Design, Implementation, and Outcomes of Using Contrasts in Learning

Abstract

This study explores variants of the Inventing-with-Contrasting-Cases (ICC) instructional model, in which students invent a representation of an underlying structure that runs throughout a set of contrasting examples. This research has two broad purposes: (1) to investigate instructional methods that may be effective for teaching general principles in science and (2) to examine the conditions under which contrasts lead learners to notice and understand the deep structures that underlie many concepts. In prior research, ICC showed strong advantages compared to more traditional instruction; students who learned via ICC methods gained a deeper understanding of the underlying structure of several physics problems (Schwartz, Chase, Oppezzo, & Chin, in press). The current research attempts to identify key ingredients for that success. We tested three ingredients to see if their combination yielded the best learning: (1) contrasting cases should highlight critical deep features of the underlying structure, (2) students need to compare and contrast the cases, and (3) students must drive towards a single, general explanation that will work for all the cases.

The study took place during a single 50-minute recitation section of an undergraduate physics course. Participants were 316 students from 11 sections of the course, and the topic of instruction was electromagnetic induction. Sections were randomly assigned to one of four conditions. Three conditions received contrasting cases. In the POE condition, students followed the predict-observe-explain method (White & Gunstone, 1992), in which they explained each case separately. In the CC condition, students simply compared and contrasted the cases. In the ICC condition, students had to invent a general explanation for all of the cases. Finally, in the ILC condition, students invented a general explanation for "low-contrast" cases.

Throughout the session, students worked in groups of 2-3 to complete a worksheet containing three contrasting cases, each one a different experimental set-up of a magnet moving in relation to a coil. Contrasts were designed to highlight the vector nature of magnetic flux. Students were instructed to recreate, test, observe, and explain the outcomes of the contrasting cases in an interactive simulation. Afterwards, students worked alone to complete a 5-question posttest which required application of the concepts to novel situations.

Initial findings suggest that simply comparing and contrasting cases is not enough to produce robust learning. To gain a deep understanding of underlying structure, students must drive towards a broad explanation or principle that can be generalized to new situations. Finally, students need “good” contrasts that sufficiently highlight critical features of that underlying structure. In addition to informing the design of science instruction, we hope this work will contribute to the broader theory of learning through contrast.

Authors

• Catherine C. Chase, Carnegie Mellon University

• Jonathan Todd Shemwell, University of Maine