AOAC ERP MICRO AUGUST 2018

Table 4-8: Delete the RSD columns. When log-transforming the data before calculation, the standard deviations do not vary with the levels, and hence RSD is of no relevance. In the calculation it is good to use many significant digits; however, in the presentation of the results, two decimals are adequate. Include units. Corrections of the values for mean difference, i.e. if the mean values are correct (please check raw data): Table 4: • Infant formula cronobacter medium level: 0.254 should be 0.03 • Sponge samples low level: 0.059 should be 0.03 Table 5: • Sponge samples medium level: - 0.051 should be -0.02 • Chicken samples high level: 0.017 should be 0.00 Table 6: • Non-fat dry milk high level: 0.100 should be -0.90 (the mean value of the candidate method is probably 4.99 instead of 3.99) • Sponge samples low level: 0.093 should be -0.45 Consider making the uninoculation levels as first rows in harmony with the other tables. Table 9: For uninoculated samples, use <0.1 log cfu/g for the mean value and na (not applicable) for sr and sR, not 0.000. Below Table 9: Please consider reformulate the sentence “ d ADifference of means <0.5 indicates no statistical significant difference between methods” to something like “ d The difference of the means should between -0.5 and +0.5 log 10 cfu/g(mL)”. The reason is that this is an agreed acceptance criteria (according to ISO 16140-2) not a statistical significance test. As paired statistical analysis is used for calculating upper and lower levels for the differences of the means and the standard deviations of the mean differences are not given in the tables, it is not possible to check the calculations of the LCL and UCL. The results are satisfactory; the precision (repeatability and reproducibility) for all levels, including low levels are below 0.5 log cfu/g and the difference of the means, except for the non-fat dry milk high level (which is probably a typing error), are satisfactory. Table 7: Correction mean difference, chicken high level 0.021 should be 0.012