Submitted to: Invertebrate Pathology International Colloquium Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: May 15, 2006
Publication Date: August 27, 2006
Citation: Wraight, S.P. 2006. Analysis, interpretation, and avoidance of difficult data in bioassay [abstract]. Proceedings of the 9th International Colloquium on Invertebrate Pathology and Microbial Control. Society for Invertebrate Pathology. 9:96. Technical Abstract: Problem data in biological assay of insect pathogens results from a number of difficult experimental design challenges, including: 1) screening of large numbers of pathogen isolates, 2) limited availability of inocula or test animals, 3) instability of pathogen preparations, 4) high variability in susceptibility of test animals, 5) censoring of data, and 6) high and variable levels of control mortality. Difficulties are exacerbated by inherent properties of pathogen dosage-mortality responses, especially low regression coefficients (slopes), imprecision of extreme lethal doses (e.g., LD95), slow development of disease, and non-normal distribution of lethal dose estimates. This presentation will examine a number of these issues and describe analytical approaches, especially for post hoc means comparisons, representing rigorous, yet powerful, alternatives to the methods commonly encountered in the present-day literature.