While many researchers have discussed how misconceptions about sampling distributions can be challenged using simulation, we have found but two warnings about how conceptual difficulties can arise or be reinforced through the use of simulation. Hodgson and Burke (2000, p. 94) found that a computer simulation of the SDM resulted in 6 of their 18 students believing that “one must draw multiple samples in order to make valid statistical inferences.” Hesterberg (1998) warns that simulations should have a large number of replications or else students “may have trouble distinguishing randomness due to random selection of data from randomness due to using small numbers of replications.”