Analysis of Students’ Feedback on C++ Programming Assessment Using an Online Survey

 Educational data mining can be used in mining student’s opinion for the enhancement of learning outcomes, instruction and learning effectiveness. Based on these reasons, this study investigated the opinions of 146 first-year student of School of Computing, Universiti Teknologi Malaysia. on learning C++ programming based on skilled-based test using an online survey. The study will also investigate some issues on the zero-score awarded to students for unsuccessfully compiled program for the skilled-based. An online survey was used to collect student opinions as data because it is simple, subject-based and easy access to the students with the help of WhatsApp groups commonly used by the students for social interaction among them and their course lecturers. The result of mining students’ opinions on learning C++ programming shows that both skill-based test 1 and 2 is found to difficult or complicated. Students’ emotions on both skill-based test 1 and 2 is found to be scary. In terms of performance, the result is slightly different with respect to program compilation. The number of students with successfully compiled programs regarding skilled-based test 1 is higher than those with otherwise. So also, the number of students with successfully compiled program regarding skilled-based test 2 is higher than those with uncompiled program. In both cases, the percentage pass is greater than 50% and failure is lower than 37%. The overall result shows that Skill-based test is difficult and it, therefore, gives scary emotions to the first-year undergraduate students of School of Computing, Universiti Teknologi, Malaysia.