Statistical Modelling 6 (2006), 175–185
A hierarchical Bayesian model for a variability
analysis of measurements of occupational n-hexane
exposure in Italy
Simona Toti
Department of Statistics ‘G. Parenti’,
University of Florence,
Viale Morgagni 59,
I–50134 Firenze
Italy
eMmail: toti@ds.unifi.it
Annibale Biggeri
Department of Statistics 'G. Parenti', University of Florence
and
Biostatistics Unit, CSPO
Florence, Italy
Alberto Baldasseroni
SA di Epidemiologia, ASF
Florence, Italy
Abstract:
This study evaluates changes over time in occupational
exposure to n-hexane by longitudinal
repeated measurements analysis of data from the
Biological Monitoring Registry from 1991 to 1998. The
main sources of variability in n-hexane exposure
among manufacturing workers in Florence province
(Italy) are inspected. The 2,5-hexanedione
concentrations in urine of industrial workers are explained
by structural, individual and factory information.
Here we analyse the effectiveness of a 1994 law on
workplace conditions based on variability decomposition
of measured 2,5-hexanedione concentrations.
We propose a hierarchical Bayesian model which takes
into account the different levels of aggregation of
data. The results show that for leather and shoe
factories, the within-subject and within-factory variance
components remain the most important over the time
of study, whereas the between-factory components
decreased in accordance with the expected effect of
the new legislation.
Keywords:
hierarchical Bayesian model; measurements variability;
occupational exposure
Downloads:
Example data
in zipped archive
back