The effect of Peer Instruction on students' construction of conceptual understanding in thermodynamics

Abstract

Many engineering classes emphasize student problem-solving skills almost to the exclusion of the understanding of underlying concepts. However, with this type of instruction, students are better rewarded by rote learning than by conceptual understanding. However, it has also been shown that the lack of conceptual understanding severely restricts the students' ability to solve a new problem since they do not have the functional understanding to use their knowledge in new situations. Alternatively, learning with understanding makes new learning easier, and leads to the development of expertise. Peer Instruction is a structured questioning process that actively involves all students in the class. In Peer Instruction, an instructor presents a multiple-choice conceptual question to the class. Students answer individually at first and next are shown a "poll" of the class responses. They then form groups and discuss the problem with peers, and finally answer again individually. Peer instruction encourages students to reflect on the problem and think through the arguments being developed and put them into their own words. Just as importantly, it provides both student and instructor with feedback regarding student understanding of the concept. This study uses the Web-based Interactive Science and Engineering (WISE) Learning Tool as a platform to investigate the effectiveness of Peer Instruction on the explicit understanding of undergraduate students in chemical engineering thermodynamics. WISE is designed to utilize the college's Wireless Laptop Initiative so that every student in a class is simultaneously engaged. The cohort in this study was 64 students in the second term of a junior level chemical engineering thermodynamics class. No rationalizations were provided for answer choices; instead, each time students answered, they were asked to reflect on their choice and provide a short written explanation. They were also asked to indicate their confidence in the answer chosen. The student explanations, both before and after group discussion, were coded based on completeness and correctness. Four question pairs were analyzed, two in which a majority of the class initially reported the correct multiple choice answer and two in which the minority had the correct answer. On exercises in which the popular answer is also the correct answer, analysis of code gains compared to code losses indicate a statistically significant increase in apparent understanding. However, on those exercises in which the majority of the class chose incorrectly, there was no statistically significant improvement of understanding for the class as whole. Furthermore, in these cases there were a significant number of students that changed their answer to the popular, incorrect answer. These results lend support to the theory of social constructivism and the consensuality principle, and can help guide the use of Peer Instruction in the classroom. © American Society for Engineering Education, 2010.