Divisions of Skills within OSS Projects

Giuri, Ploner, Rullani, and Torrisi (2010) analyze the talent, groupings, and performance of open source software (OSS) projects within Sourceforge.net. The hypothesis relevant to my research pertains to the project performance affected by the modularity of tasks within the project. Giuri, Ploner, Rullani, and Torrisi (2010) did not have access to the source code, so the modularity does not concern the code, but rather the concept of the project being divided into several subprojects. Additionally, the authors focused on the activity within each of the subprojects and the activity’s affect on the whole project. Activity is defined as a fixed bug, patch, or new enhancement developed through a request and released in an update, product release, or CVS (concurrent version system) commitment.

They concluded that the organizational modularity of the project does have a positive effect on the project; however, the outcome of modularity to performance is nonlinear. Additional support to project performance arrives through developer skill level. Across the membership, the standard deviation of the experience level and skills remain significant when the equation includes controls. Moreover, the data points to better performance with segregation among member skill levels. The project profits from grouping specialists, generalists, high level members and low level members separately.

On the other hand, the study also suggests a negative consequence to interaction among skill diversity and modularity at the project level. The results lead to the concept that two organizational approaches exist: a “workshop-like” and “factory-like” organization. The workshop arena relies on its members with diversified skills who multitask; whereas, the factory concept involves high modularity and relatively low level skill diversity.

Reference

Giuri, P., Ploner, M., Rullani, F., & Torrisi, S. (2010). Skills, division of labor and performance in collective inventions: Evidence from open source software. International Journal of Industrial Organization, 28(1), 54-68.