Further on creating normal density functions in Microsoft® Excel

Simulations play an increasing role in metrology and analytical chemistry, particularly the use of simulated normal distributions. Microsoft Excel is available to almost everybody and normal distributions can be simulated using either an innate function or other algorithms. We explore the success of five different algorithms to simulate normal distributions and compare the outcome with a simulation coded in R. The normality of 106 data points simulated by the chosen algorithms was tested by different recognized statistical procedures and Q–Q plots. They all failed the statistical procedures, but evaluating the Q–Q plots revealed different types of deviations. It also showed that the deviations, although statistically significant, most likely do not have any practical relevance in laboratory work.