Search code examples
statisticsnetlogoagent-based-modelingbehaviorspace

How much repetitions of virtual experiments are enough to be "statistically sure" about its findings?


After reading this article about BehaviourSpace in Netlogo in which the author stated that "[...] the model is stochastic, you need to run each parameter setting many times to be statistically sure about this", I was wondering how much repetitions are enough to be "statistically sure" about the findings in virtual experiments?


Solution

  • In general you can use the formula mentioned in the question here to establish confidence in a randomized (monte carlo) simulation: https://stats.stackexchange.com/questions/95779/required-number-of-simulations-for-monte-carlo-analysis