Supporting Students in C++ Programming Courses with Automatic Program Style Assessment

Abstract

Professional programmers need common coding conventions to assure co-operation and a degree of quality of the software. Novice programmers, however, easily forget issues of programming style in their programming coursework. In particular with large classes, students may pass several courses without learning elements of programming style. This is often due to shortage of tutor work to give students thorough feedback on their coursework. Incorporating issues of style into programming courses is too often neglected and students are hoped to learn these issues by themselves. To deal with the problem, a set of coding rules was collected and justified to be used in C++ programming courses of the university. An automatic C++ programming style analyzer tool was implemented to ensure that students were following the rules. Students can freely use this tool to improve the quality of their coursework, and tutors can use it for assessing the assignments. Assessment rules and criteria can be easily adjusted according to the needs of the course or the exercise in question. Since basic programming style issues are assessed by the students independently before coursework submission, the teaching staff can concentrate on giving feedback on the more advanced features of program design and course specific issues. The approach seems to have tackled the basic problem well. The students learn to pay better attention to their coding practices and they develop themselves good basic programming habits already on their first courses. After the first shock of having more requirements for the practical programming assignments, students have regarded the automatic style analysis as a useful aid in the programming courses. The quality of coursework has improved, and including a systematic assessment of programming style also improved the partitioning of contents and design of the courses involved. With more careful monitoring and evaluation of the tool usage, we hope to gain even better results in improving the understanding and adoption of good coding principles as part of students' personal programming practices. (Contains 8 figures and 1 table.)