An Introduction to Statistical Modelling. Wojtele J. Krzanowski, Arnold, London, 1998. No. of pages: x+252. Price: £19.99. ISBN 0-340-69185-9

✍ Scribed by Barbara McKnight

• Book ID: 101239722
• Publisher: John Wiley and Sons
• Year: 2000
• Tongue: English
• Weight: 48 KB
• Volume: 19
• Category: Article
• ISSN: 0277-6715
• DOI: 10.1002/(sici)1097-0258(20000530)19:10<1380::aid-sim401>3.0.co;2-q

✦ Synopsis

a small increase in relative risk will result in a large increase in absolute risk. Whatever the size of the risk and however uncertain, we do agree that no one should involuntarily be exposed to ETS.

In a rather technical review, A. C. Heath, P. A. F. Madden and N. G. Martin illustrate the statistical methods that have been used to test for genetic in#uences on smoking behaviour. Using data from three large twin panel studies conducted in Finland, Australia and the U.S.A. they focus on two complementary approaches: a genetic model-"tting, and logistic regression models for the signi"cance of genetic e!ects of smoking. The authors conclude that there is a signi"cant and substantial genetic in#uence on smoking persistence and that genetic e!ects cannot be accounted for by the inheritance of sociodemographic or personality variables. We would feel more comfortable with this conclusion if it were corroborated by adoption studies.

In a short "nal review, J. Stapleton addresses recent British trends in prevalence and cessation, tobacco dependence and the nicotine withdrawal