The heterogeneity associated with the spatial distribution of organisms is an awkward problem in ecology because this heterogeneity directly depends on the sampling scale. To specify the scope of the influence of sampling scale on the level of species aggregation, we need data sets that entail excessive sampling costs in situ. To find a solution for this problem, we can use models to simulate patterns of organisms. These models are often very complex models that take into account heterogeneity of habitats and displacement or longevity of studied species. In this article, we introduce a new stochastic model to simulate patterns for one taxon and we want this model to be parsimonious, i.e., with few parameters and able to simulate observed patterns. This model is based on an aggregation–repulsion rule. This aggregation–repulsion rule is defined by two parameters. On a large scale, the number of aggregates present on the pattern is the first parameter. On a smaller scale, the level of aggregation–repulsion among individuals is determined by a probability distribution. These two parameters are estimated from field data set in a robust way so that the simulated patterns reflect the observed heterogeneity. We apply this model to entomological data: four Diptera families, namely the Sciaridae, Phoridae, Cecidomyiidae, and Empididae. The field data for the Phoridae family are used to simulate sampling using different trap sizes. We record changes in the coefficient of variation (C) as a function of the sampling scale, and we can suggest to ecologists emergence traps of 0.6 m2, in other words a square 77 × 77 cm trap, to obtain a C value under 20%.