PRACTICABLE APPROACHES FOR THE CHARACTERI ZATION OF VARIANCE

Like all analytical methods, the standard plate count in food contains errors to a certain extent. While the precision of plate count result s can be estimated experimentally, it is not possible to determine the ir accuracy by means of experiments. The review presented gives an acc ount of practicable approaches for the characterization of variance an d its classification into different components. The investigation show s that variance between parallel samples or parallel laboratories resp ectively constitutes the biggest part of variance while methodological effects remain comparatively insignificant. Due to the lack of a gene rally established model for the classification of variance-components, it is difficult to compare data that have been published up to now.