Despite the efforts to achieve high quality data and to use efficient statistical estimators, the results of the NFI are not free of errors. A complete census of the Swiss forests is impossible due to the necessary costs, the available personnel, and the required time from the beginning of the survey up to and including the publication of the results; thus no other alternative to a sample based approach for the NFI exists. In a sample based survey a small portion (sample) is selected from the entire population of the Swiss forests. The selected elements (trees or sample plots) are precisely assessed. The elements included in the sample are then used to draw an inference about the entire population. Inferring about the entire population (e.g., Swiss forests) based on the sample has its roots in theoretical probability assumptions. Simply spoken, the probabilities with which the individual population elements are selected for the sample are taken into account during the derivation of the statistical parameters (e.g., mean values, proportions, and totals). For the inference, the statistical parameters that were calculated using the sample data are applied to the entire population.