 |
 |
hello all, I wonder if someone could solve this basic question for me: If one has a collection of cancer patient data, and age is amongst this data, people tend to normalise the data for age. If, for example, there is data on urinary symptoms as well and one wants to see if these symptoms are associated with disease, it is usual to perform logistic regression (disease is a binary outcome) in which one adjusts for age and maybe other variables as well. However, since age will be adjusted for in the regression, is it necessary to normalise the data for age before the regression analysis? I would have thought not, but would like some comments from you as well. Many thanks