Automatic Removal of Outliers from Regression and GLMs
A well-known problem with linear regression, binary logit, ordered logit, and other GLMs, is that a small number of rogue observations can cause the results to be misleading. For example, with data on income, where people are meant to write their income in dollars, maybe one person writes their income as 50, meaning $50,000, and […]
5 Alternatives to the Default R Outputs for GLMs and Linear Models
The standard summary outputs from the glm and lm summary methods are a case in point. If you have been using R for as long as I have (19 or 20 years…) you will no doubt have a certain affection for them, but to a new user they are both ugly and not optimized to […]
