Abstract:
|
Teaching students new statistical programming languages while simultaneously teaching them how to debug erroneous code is challenging. Having learned only correct code in class, students cannot debug foreign errors they generate out of class. We need to bridge the gap between error-free learning in class and error-full learning out of class by balancing error-free and error-full teaching. We conducted a prospective pretest-posttest pilot study of n=18 Milken Institute SPH graduate students who voluntarily attended a Debugging SAS Seminar held weekly from Sept-Nov 2018. Each seminar had a 10-min error-free lecture, 15-min error-free assignment, 5-min break, 10-min error-full lecture, and 15-min error-full assignment. Student performance and preference were compared using binomial tests. While 22.2% (n=4) successfully completed both assignments and 55.6% (n=10) completed neither, 5.6% (n=1) successfully completed only the error-free assignment and 16.7% (n=3) successfully completed only the error-full assignment (p=0.63). Significantly more students 80% (n=12) preferred error-full to error-free teaching (p=0.035). We will evaluate error-full teaching in an introductory SAS course.
|