ABSTRACT
Generalized linear models (GLMs) are a lot like these early mechanical computers. The moving pieces within them, the parameters, interact to produce non-obvious predictions. The most common and useful generalized linear models are models for counts. Binomial GLMs are appropriate when the outcome is a count from zero to some known upper bound. The island societies of Oceania provide a natural experiment in technological evolution. Different historical island populations possessed tool kits of different size. Poisson regression is a GLM that models a count with an unknown maximum—number of elephants in Kenya, number of applications to a PhD program, number of significance tests in an issue of Psychological Science.
