Recently, I was wondering if there has been any meaningful work done to apply linear regression analysis to determine the correct balance between various dependent variables used in an Agile process. For example, what is the impact of architectural runway, number of developers, hours per day, length of an iteration etc. on the quality and output of a software project? There is no question many ideas and opinions exists on the right value for these variables. However, I would like to know if there is “statistical proof” on how each of these variables impacts a software project? If anyone out there has a work they I can reference, I would appreciate it?