PPIG 2014 - 25th Annual Workshop
Affective Learning with Online Software Tutors for Programming
Amruth N. Kumar
Abstract: We conducted a study to see if there was any difference in the affective learning of students who used software tutors to learn programming concepts. We used two software tutors – on arithmetic expression evaluation and tracing selection statements for this study. Data was collected over five semesters with the arithmetic tutor and over four semesters with the selection tutor, yielding a sample size in the thousands. The data was analysed using ANOVA, with sex (male versus female), representation (Caucasians and Asians versus other races), and discipline (Computer Science versus other disciplines) as fixed factors. We found no difference in the affective learning of any demographic group - all the groups felt in equal measure that they had learned using the software tutors. This is a positive result since affective learning complements cognitive learning. But, we did find statistically significant difference between the sexes and between the representation groups on prior-preparedness on arithmetic tutor, but not selection tutor. This may be because arithmetic tutor covered concepts that students had been exposed to in high school, but the concepts covered by selection tutor were unique to programming and unlikely to have been seen by students who had had no prior programming experience. Finally, non-Computer Science majors learned significantly more concepts than Computer Science majors on both the tutors, even though no significant difference was found between the two disciplinary groups on prior-preparedness. This may allude to difference in perceived self-efficacy of the two groups towards learning from software tutors.