Effect of Pipette Type and Sample Size on Plate Count Results

Abstract

Summary Raw and pasteurized milk samples were plated by seven analysts using 0.1- and 1.0-ml quantities (of appropriate dilutions) dispensed by 0.1-ml glass, 1.1-ml plastic and 1.1- and 2.2-ml glass pipettes. Incubation was at 32 C for 48 hr. Data were subjected to an analysis of variance and the standard deviation as a ratio was calculated. The procedures were compared from two points of view, first to determine if biases existed between the different methods and second if the techniques differed in their random error. No appreciable differences were found between types of pipettes when pasteurized milk was plated. The quantity used, however, was important. Use of a 0.1-ml sample size was associated with more variable results which averaged 10% higher than when a 1.0-ml quantity was plated. Statistical analysis of test results obtained with pasteurized milk suggests: (a) when 0.1-ml samples were used least variation was encountered with the 0.1-ml glass pipette and most variation with the 1.1-ml glass pipette, and (b) when 1.0-ml samples were used very little difference appeared between pipettes although the 1.1-ml plastic pipette tended toward less variation. When raw milk contained large numbers of bacteria, the plastic pipette tended to give higher results than glass pipettes when the 1.0-ml quantity was used; however, these differences with raw milk were not found to be statistically significant. Standard deviations as a ratio revealed: (a) greatest variation with the 2.2-ml glass pipette and least with a 0.1-ml glass or 1.1-ml plastic pipette when a 0.1-ml quantity was used, and (b) greatest variation with a 1.1-ml plastic pipette and least with a 2.2-ml glass pipette when a 1.0-ml sample size was used.