Using machine tools to analyze changes in students' ethical thinking

Engineering ethics education entails the development of the ability to recognize the social, cultural, environmental, and global implications of engineering practice. Instructional activities often involve discourse among students and verbal and written responses to ethical issues and dilemmas. The present research applies two machine methods to extract the dominant concepts in engineering undergraduates' essays that were written at the beginning and end of an engineering ethics course. The two methods were Linguistic Inquiry and Word Count (LIWC) and naïve Bayesian analysis. Both methods showed little overall change in the conceptual basis of beginning versus end of semester essays. Closer analyses of the Bayesian results suggested there were observable individual differences in the essays. Further analysis of these differences may aid in better understanding which students changed their ethical thinking from the beginning to the end of the course. In the Discussion we suggest several ways in which success with these machine methods could aid instruction.