2021 Volume 25 Issue 1 Pages 51-60
Evaluation of regression models is essential for identifying avoidable errors in the data manipulation process and model building, and assessing the robustness of model results. The model evaluation has two steps; assessment of the overall goodness of fit and identifying influential observations. This article covers a fundamental concept of the model evaluation and provides a few strategies to deal with poorly fitting models. A tutorial material using R is developed where readers can replicate line-by-line a model building and evaluation process.