In this paper, we present a general methodology to assess the performance of an ad-hoc network equipped with a proactive routing protocol, namely the Open Link State Routing protocol. We are interested in identifying the parameters which strongly affect the overall performance. To this aim, and contrary to the common use of computing means and variances only, we use a hypothesis test-based toolkit, because of the large number of factors which interact with our testbed. We discovered that there is a hop-count threshold, after which the performance deteriorates. For instance, the statistical analysis has shown that after the hopcount threshold the variability of our data increases, i.e. a marked oscillatory behaviour arises. The results of this work confirm the need of adaptive selection of the routing protocol parameters.