Agreement between dyeing methods. The average value of the number of cells per field (x-ijk. or ȳijk.) for each PC filter was determined by the NB model, which is adapted to the number of cells in the 12 fields. ANOVA models with log-transformed values of x-ijk. or ȳijk. satisfied the assumptions of normality and homoscedacity, while different transformations of xijkl and yijkl did not reveal a normal model for xijkl and yijkl, mainly due to the fact that the accidental error in equation 1 was a NB distribution. This result implies that the accidental error due to subsampling, estimated from Triplicate subsamples, could be analyzed as log-normal distributions. We can conclude that the weight of the apple pouch is 5 lbs ± 8%. Think about how this percentage of uncertainty would change if the bag of apples were half as heavy, but the uncertainty in the weight remained the same.

Note for future calculations: When calculating the percentage of uncertainty, always think about the need to multiply the share by 100%. If you don`t, you have a decimal amount, no percentage. Estimated accuracy. In the statistical sense of the term, the accuracy of the measurement is the proximity of a measure to its actual value (56). In estimating the accuracy of the number of bacteria by microscopic methods, it is difficult to obtain the actual value. Chae et al. (13) considered the actual value as the best available measure for true concentration, and this concept led them to search for the counts obtained under the most favourable conditions as the true estimates of the plethora of particles under the microscope. In the practice of the concept, the authors used the larger counts with the higher density of particles per field than the actual number of particles per field.

In the Lisle et al. (36) study, the actual values were nominally determined by the preparation of samples with dilutions from a single high-density stock, whose cell abundance was enumerated by an interlaborative standardized growth curve of a bacterial strain. Since stock frequency estimates and dilution operations are subject to accidental errors, we believe that Lisle et al.`s approach is only acceptable if accuracy is interpreted on a relative scale. Chae et al.`s approach will likely underestimate accuracy, as the highest particle density may be an overestimation or near the upper limit of possible random variations, instead of the average.