A fresh look at novice programmers' performance and their teachers' expectations.

ABSTRACTThis paper describes the results of an ITiCSE working group convened in 2013 to review and revisit the influential ITiCSE 2001 McCracken working group that reported [18] on novice programmers' ability to solve a specified programming problem. Like that study, the one described here asked students to implement a simple program. Unlike the original study, students' in this study were given significant scaffolding for their efforts, including a test harness. Their knowledge of programming concepts was also assessed via a standard language-neutral survey. One of the significant findings of the original working group was that students were less successful at the programming task than their teachers expected, so in this study teachers' expectations were explicitly gathered and matched with students' performance. This study found a significant correlation between students' performance in the practical task and the survey, and a significant effect on performance in the practical task attributable to the use of the test harness. The study also found a much better correlation between teachers' expectations of their students' performance than in the 2001 working group.