Hey anons, can anyone who's good with data analysis - particularly Generalized Linear Models and regression in R help plz:

State the advantages of using a Poisson GLM on raw counts, compared to using a general linear model (LM)—i.e. a model with a Gaussian error term—with a transformed response variable. In your answer, list the properties of count data that can make them unsuitable for use in an LM and state the reasons why a transformation may not resolve these problems.