Designing A Statistics Course For Chemical Engineers

Abstract

The Department of Chemical Engineering at Ohio University redesigned an existing course in experimental design and statistics. The revision was motivated by assessment information from a variety of sources: course-based assessment in our senior Unit Operations laboratory, exit surveys of seniors, surveys of alumni 2 years after graduation and input from our departmental advisory board. The consensus of faculty, students, alumni, and the advisory board was that (1) a solid foundation in statistics is important preparation for industrial engineering practice as well as for advanced degree work in engineering and (2) “solid foundation” means that graduates can select and execute appropriate statistical techniques to analyze real data and interpret the results. In spite of having a statistics course in our curriculum, graduates did not leave with the solid foundation we wanted. In particular, our seniors showed unsatisfactory ability to frame a problem in terms of a hypothesis that can be tested statistically and unsatisfactory ability to select an appropriate statistical test. New graduates were only beginning to operate at the desirable higher levels of analysis, synthesis, and evaluation. As part of a strategy to address this problem, our statistics course for juniors was redesigned with input from our faculty and from industrial members of the advisory board. The new course emphasizes software rather than hand calculations, introduces application and follows up with theory, and uses case studies from industry and from academic research. This course is not isolated in our curriculum. Statistical analysis is now a required part of projects in Heat Transfer and Kinetics, and continues to be emphasized in Unit Operations. In this talk, we reveal the motivation for emphasizing statistics in our curriculum, the structure of the re-designed course, and the assessment methods being used to gauge student learning in this course.

Why Teach Statistics? Statistical methods of data analysis are valuable tools to chemical engineers in both research and in industrial practice. Consider this quote from a recent National Science Foundation program announcement (italics added). “Projects must use appropriate quantitative methods, and teams should include individual(s) with demonstrated expertise in the quantitative methods to be used. Quantitative methods may include: conceptual, mathematical or computational models; computer simulation; artificial intelligence techniques; hypothesis testing; statistics; visualization; or database development. Mathematical models must include estimates of uncertainty, and experiments should assess power and precision.” “Six-Sigma”, the currently popular industrial philosophy and method for quality control, is based on statistical methods of process analysis and decision-making. Articles about “Six-Sigma” have recently featured in Chemical Engineering Progress.1,2 Every engineer is expected to participate in Six-Sigma process improvement, not just those assigned to a process or quality control department.

Proceedings of the 2003 American Society for Engineering Education Annual Conference & Exposition Copyright © 2003, American Society for Engineering Education