Coding Categorical Variables

Описание: In BIOS 6611 we have thus far worked primarily with binary categorical variables using indicator variables. When we have categorical variables with 3 or groups we can also provide indicator variables structured to this setting. More generally, our use of indicator variables in a regression model that includes the intercept is known as the reference cell model.

Alternative strategies exist for coding categorical variables. We briefly touch on effect coding, cell means, and treating the variable as continuous (which carries very, very strong assumptions).

A video for the Biostatistical Methods I (BIOS 6611) course in the Department of Biostatistics and Informatics at the University of Colorado-Anschutz Medical Campus taught by Dr. Alex Kaizer. Slides and additional material available at https://www.alexkaizer.com/bios_6611.

Table of Contents:

00:00 - Intro Song
00:15 - Welcome
00:37 - Variables with More Than 2 Categories
01:41 - Indicator ("Dummy") Variables
03:13 - Reference Cell Models
08:39 - Change the Reference Category
10:50 - Reference Cell Interpretation
14:09 - Other Coding Schemes
15:30 - Effect Coding
16:45 - Cell Means Model
18:47 - Treat It Continuous (CAUTION!)