| This exploratory study considers choice theory, decision theory, and the constructivist theory of education to explore college-level computer science learners' behavior when presented with multiple instructional modes (instructional methods for the presentation of course content, such as video, text, audio, animation, etc.) in an online learning environment. Demographic, performance, and behavior data were analyzed using exploratory data analysis techniques--including recursive partitioning, linear modeling, analysis of variance, and elemental graphics visualizations--to determine which factors serve as the best predictors of learner behavior, i.e., which factors best predict learners' choices of instructional modes. Learners enrolled in introduction to computer science courses at 4 two- or four-year colleges in the Northeastern United States completed 3 online learning modules. Each module included 5 different instructional modes learners could choose to access. Participant behaviors while completing the modules were recorded using learning management system (LMS) logs. The study found that participants did not avail themselves of the multiple instructional modes presented in the online modules. Demographic variables such as age, gender, number of semesters in college, and previous online course experience were not found to be consistent, statistically significant predictors of participants' behavior. |
