Bootstrap inference on the Poisson rates for grouped data

Research Report SRR97-001

Bootstrap inference on the Poisson rates for grouped data

Kee-Won Lee

Abstract: We present how bootstrap methods can be used to conduct inference on the rates of Poisson distributions when only the grouped data are available. A data set obtained from a pathology laboratory test is used for illustration. Traditional asymptotic methods are compared with bootstrap methods in computing the estimated standard errors and achieved significance levels for one sample and two sample tests. Bootstrap methods are shown to possess a definite advantage over the asymptotic methods in that small sample distribution of the relevant statistic can be readily obtained from the bootstrap copies.

This service is maintained by the Mathematical Sciences Institute (MSI)
Comments to webmaster@maths.anu.edu.au URL: http://wwwmaths.anu.edu.au/