Implementing flexible hybrid instruction in an electrical engineering course: The best of three worlds?

•The study explores a hybrid format designed for more customized learning.•Class attendance only predictes performance when course material is difficult.•Time management is a serious issue when students are given too much liberty.•Motivation predicates exam performance, especially when the content is challenging.•The success of this hybrid format depends on student skills and course difficulty.

This study explored a modified version of hybrid instruction, referred to as the flexible hybrid format, in a lower division electrical engineering course offered at a large public university. The objective of the study is to use longitudinal data to investigate the impact of class attendance, out-of-class study time, and motivation on student exam performance. Generalized least squares and fixed effects models were used in the analyses. It was found that class attendance was indispensable; it was associated with exam performance even when all essential course material was made available online and students generally rated the online instruction component to be of higher quality. The benefit of class attendance was then explained by the ICAP hypothesis and spaced learning practice and it was suggested that online education might be more effective in teaching relatively simpler contents. Out-of-class effort significantly predicated performance in previous weeks, but not in the final period. The harmful effect of cramming was cited to explain this phenomenon. Hence, by implication, time management might be an issue in a flexible hybrid environment. Finally, motivation was found to be a robust predicator of performance and its effect was the strongest when the course was at its most challenging stage. Besides, the relationship between motivation and exam performance was likely to be bidirectional, as higher motivation resulted in better performance, which in turn further boosted motivation. Based on current findings, directions for future research were also suggested to verify our claims and improve our implementation.