Zero-inflated proportion data models applied to a biological control assay

作者: A. M. C. Vieira , J. P. Hinde , C. G. B. Demetrio

DOI: 10.1080/02664760021673

关键词: Multinomial distributionNegative binomial distributionOverdispersionCount dataQuasi-likelihoodBinomial testBinomial distributionBeta-binomial distributionStatisticsMathematicsEconometrics

摘要: Biological control of pests is an important branch entomology, providing environmentally friendly forms crop protection. Bioassays are used to find the optimal conditions for production parasites and strategies application in field. In some these assays, proportions measured and, often, data have inflated number zeros. this work, six models will be applied sets obtained from biological assays Diatraea saccharalis , a common pest sugar cane production. A natural choice modelling proportion binomial model. The second model overdispersed version model, estimated by quasi-likelihood method. This was initially built overdispersion generated individual variability probability success. When interest only positive data, can based on truncated distribution its version. last two include zero proport...

参考文章(0)