In toxicology, data often deviate from normality due to the nature of the biological responses being measured. For instance, binary outcomes (e.g., survival vs. death), count data (e.g., number of adverse effects), and proportional data (e.g., percentage of affected cells) are common. GLMs accommodate these types of data by allowing the response variable to have a distribution from the exponential family, such as binomial, Poisson, or gamma distributions.
