I am part of a group which is using an internally developed tool to produce quantitative measures of software developer behavior. This tool produces a detailed log of all of the major activities that take place during a software development session. This data collection is completely unobtrusive to the software developer. In doing this we have developed a set of models and associated metrics which we are now attempting to validate in several empirical studies on novice programmers. A simple example of the type of model and metric we are attempting to validate is whether there is a link between the time elapsed from development start to the first compilation request to the overall performance of a novice programmer (measured by the score received on a software assignment). Another simple example is the measurement of software developer bias when performing the reporting of actual (versus estimated) development time (as per Humphrey's PSP). I am posting to this group to attempt to locate other research and/or existing literature in this area. While our investigations have found some small caches of literature in this area (primarily in the area of debugging), we've found little that is on point. Any help would be greatly appreciated. Mark D. Campbell [EMAIL PROTECTED]
