ABSTRACT
This is a very short chapter which begins with a Poisson regression example in this section and then turns to a brief discussion on fitting general nonlinear models in Section 9.2.
The example uses data obtained from STATLIB-DASL (https://lib.stat.cmu. edu/, hosted by CarnegieMellon). It contains cigarette consumption (a measure of the number of cigarettes smoked per capita) along with death rates (per 1000) for lung and other cancers, for 43 states and the District of Columbia. Two cases with consumption greater than 40 were eliminated for this analysis. The cigarette consumption is clearly an estimate rather than an exact value. Here we illustrate fitting a Poisson regression model with log-link relating the lung cancer rate, Y , to cigarette consumption, x, allowing for error in cigarette consumption.
