The effect of sampling variability on statistical order point computation

Recognition of sampling variablity is a fundamental issue in any statistical analysis, yet has been largely overlooked in the area of statistical order point computation. This article is an illustration that a statistically computed order point is a random variable, and will vary from sample to sample around the ‘true’ order point required to achieve some desired long-run average service level. In practice, inventory managers should be aware that statistical analysis may provide an order point for any single inventory item quite different from that associated with the desired service level.