Expertise in Debugging Computer Programs: An Analysis of the Content of Verbal Protocols

The results of an exploratory study that investigated expert and novice debugging processes with the aim of contributing to a general theory of programming expertise are reported. The research had two major objectives. First, it sought to control for the variability in individual problem solving by incorporating certain aspects of subjects' debugging processes into the debugging model. The criterion of expertise was the subjects' ability to chunk the program they were required to debug. This method proved effective at explaining much of the variability in debugging time and the number of errors subjects made and, hence, afforded information on the differences in the debugging processes of experts and novices. Second, the study assessed the nature of the differences in expert and novice debugging processes by analyzing the data at two levels, representing different degrees of dependence on the current problem solving environment. The literature suggests that novices rely on formal models in problem solving while experts react principally to the task environment as revealed during problem solving. The method used was verbal protocol analysis. Data were collected from sixteen programmers employed by the same organization. An a priori task analysis was used as the basis for a content analysis of the protocol data. The research provided some support for the notion that experts are situation-dependent problem solvers, while novices are situation-independent problem solvers.