Are Two Heads Always Better Than One in Collaborative Programming?

The software industry continues to struggle with producing quality software in the most efficient manner. Anecdotal evidence suggests that the use of newer, innovative development methods, which embrace high levels of collaboration, may be a viable solution to this problem. While collaboration has always been used, these techniques emphasize high levels of interpersonal collaboration during the entire development process. An example of collaborative programming that is gaining interest is pair programming. A relatively unexplored element of collaborative programming is how the diversity of the individuals paired in the programming dyad impacts performance outcomes. We propose an experiment with the primary emphasis on how the diversity of individual developer characteristics impacts collaborative programming results, with a focus on creating the “best” pairing possible in order to raise performance and maximize outcomes.