Examining student constructed explanations of thermodynamics using lexical analysis

Abstract

Thermodynamics can be challenging to students, thus improving thermodynamics instruction and assessment is an important area of science and engineering education research. Constructed response assessments can reveal the complexity of students' ideas about thermodynamics. We investigate the use of lexical analysis software for examining students' constructed responses using a group of three questions related to reaction thermodynamics. These questions were administered to students in a large enrollment undergraduate introductory science course and examined learning at two levels of Bloom's Taxonomy: comprehension and application. Our results show that students are able to identify correct statements about thermodynamics in a multiple choice context but fail to construct correct explanations using thermodynamic concepts. Lexical analysis revealed that students who provided correct explanations incorporated more correct concepts and/or made more connections among these concepts than did students with incorrect explanations. Lexical analysis provided insight into student understanding by revealing heterogeneous ideas that were masked in multiple choice versions of the assessment.