Packet 9: Categorical Data Analysis

Packet 9: Categorical Data Analysis

1. Contigency Tables

   • Marginal Association
     • Aggregate over third variable
       • (n11*n22)/(n12*n21)
   • Conditional Association
     • Decide yes or no for the given third variable, then only pull those values of the the other 2 variables
       • (n11*n22)/(n12*n21)

2. Relationships of Variables

   • Mutual
     • (X,Y,Z)
       • glm(count ~ X + Y + Z , family = "poisson")
   • Joint
     • (X,YZ)
       • glm(count ~ X + Y + Z + YZ , family = "poisson")
   • Conditional
     • (XY,XZ)
       • glm(count ~ X + Y + Z + XY + XZ, family = "poisson")
   • Saturated
     • (XYZ)
       • glm(count ~ X + Y + Z + YZ + XY + XZ + XYZ , family = "poisson")
   • Uniform Association
     • (XY,YZ,XZ)
       • glm(count ~ X + Y + Z + YZ + XY + XZ, family = "poisson")

3. Joshua Chapple, Tommy Huber, Billy Hornby, Mariana Suarez