In his last editorial forSIGCOMM's Computer Communication Review, S. Keshav explains his frustation in finding that among the 30 papers he reviewed for SIGCOMM, most contained flaws in their statistical methodology. The networking community often uses experimements and simulations, but I agree with Keshav that many papers often neglect to include confidence intervals and other statistical information in measurements and simulation results.

We can hope that Keshav's editorial will lead reviewers to better check the statistical value of the submitted papers. Keshav's also suggests that PhD students should read books or online resources on statistics or follow courses on this subject. Although a bit dated, Raj Jain's The Art of Computer Systems Performance Analysis: Techniques for Experimental Design, Measurement, Simulation, and Modeling is still a recommended book for network researchers.