Statistical analysis of asbestos fibre counts

Gary W. Oehlert, Richard J. Lee, Drew Van Orden

Research output: Contribution to journalArticlepeer-review

27 Scopus citations

Abstract

We present methods for estimating asbestos fibre concentrations based on the negative binomial distribution for fibre counts. This distribution is preferred to the usual Poisson due to the extra variability encountered in the data. We extend the method to include estimation of the parameters of a log‐normal population of concentrations, and deal with the problem of zero counts. We illustrate the methods on a set of 921 fibre counts from 177 school buildings.

Original languageEnglish (US)
Pages (from-to)115-126
Number of pages12
JournalEnvironmetrics
Volume6
Issue number2
DOIs
StatePublished - Jan 1 1995

Keywords

  • EM algorithm
  • asbestos
  • maximum likelihood
  • negative binomial
  • zero counts

Fingerprint Dive into the research topics of 'Statistical analysis of asbestos fibre counts'. Together they form a unique fingerprint.

Cite this