AWWA WQTC69417

The objectives of this study were to: enable quantitative estimation of variability in microbial data as a technique for evaluating experimental design; and, provide a tool that enables comparison of sampling and methodological errors so that analysts can focus efforts on reducing the dominant error. Variability in enumeration methods: five categories of sampling and methodological errors have been identified in probabilistic models that enable more rigorous statistical analysis of microbial enumeration data, including the following: Representative sampling - variability due to spatial or temporal heterogeneity of the source (samples are not replicates); Random sampling error - variability in the number of microbial particles captured in replicate samples of the same volume; Random analytical error - variability in the number of microbial particles observed due to method losses; Non-constant analytical recovery - variability in the recovery efficiency of the method between equivalently processed samples; and, Counting error - variability in the number of observations between repeat counts of the same sample by the same or different analysts. Includes 4 references, figures.