(107h) Comparing Formative and Summative Auto-Graded Problems for a Material and Energy Balances Course | AIChE

(107h) Comparing Formative and Summative Auto-Graded Problems for a Material and Energy Balances Course

Authors 

Liberatore, M. - Presenter, University of Toledo
Yanosko, S., University of Toledo
Digital tools providing students with immediate feedback are well aligned with many best practices related to learning. Specifically, auto-graded, randomized problems for a Material and Energy Balances course has quantified correct and attempts for over 100,000 problems over more than 5 cohorts. These “big data” were obtained with using the fully interactive online textbook titled the Material and Energy Balances zyBook. In general, a median correct over 90% has been presented in previous work with median attempts before correct less than 3 for almost all problems. Thus, with unlimited attempts before weekly due dates, students are generally successful in solving auto-graded problems. On one hand, all cohorts of interest have used in-section, formative auto-graded problems that apply scaffolding, i.e., moving from easier to more difficult problems within a set of 3 to 7 problems. On the other hand, summative, end-of-chapter auto-graded problems were added and analyzed for three recent cohorts. Summative problems include both traditional problems as well as student-written YouTube problems, which has been discussed and published in recent years. Research questions included: 1. Do students correctly solve formative problems more readily than summative problems?, 2. Do formative and summative auto-graded problems correlate differently with final course grades?, and 3. Can too many problems be assigned? Some discussion on variation and reproducibility from cohort to cohort will also be discussed.