By Andrew Sage and Ulrike Genschel (Iowa State University)
We investigate the impact of using real data with applications in engineering on the perceived relevance of statistical inference among students enrolled in an engineering statistics course at Iowa State University. Students completed a survey measuring attitude and engagement with course material (Floyd, Harrington, Santiago, 2009) before and after participating in a group project involving the use of complex, non-textbook data in an engineering context. In the project, students applied multiple regression to model the strength of concrete using the dataset described by Yeh (1998), which is available in the UCI machine learning repository. The class, taught through the statistics department, consists of 43 (mostly industrial) engineering majors, the majority of whom are sophomores and juniors. The results of the project will have implications on future instructional strategies for this course as well as other undergraduate statistical courses at Iowa State and other institutions.