Educational Technology & Society

This study investigated the instructional efficiency of adaptive instruction to learner expertise in the domain of C programming language with college students. It also aimed to investigate whether a bimodal process-oriented worked-out example (WOE) could effectively control extraneous cognitive load and further improve instructional efficiency. For this purpose, a learner-paced problem-solving e-learning environment was developed. A total of 112 college students participated and they were randomly divided into four groups (adaptive and bimodal, adaptive and unimodal, fixed and bimodal, and fixed and unimodal) when they logged into the problem-solving e-learning environment. After removing uncompleted or repeated data, data from 96 students were used for a series of ANOVA and ANCOVA. The findings showed that the adaptive instruction groups showed significantly higher instructional efficiency than the fixed instruction groups. Although there was no significant difference between the bimodal and the unimodal WOE groups, the bimodal WOE groups showed lower mental effort, higher knowledge acquisition, and instructional efficiency. In addition, the bimodal WOE condition was more efficient in both adaptive and fixed conditions. Based on these findings, it was concluded that the adaptive instruction method and the bimodal process-oriented WOE effectively controlled cognitive load and thereby successfully and efficiently led to schema construction and automation. The insight gained through this study may inform instructional designers seeking to enhance their understanding of efficient instructional design within the cognitive load theory framework.