GLM Model

Question 1

Types of Data

Answer 1

Numbers: Age/Score = Continuous

Groups: Gender/Condition = Categorical

Question 2

Bayesian v. Frequentist

Answer 2

Frequentist: Traditional Stats (Like p-values)

Bayesian: Newer; focuses on probability

Question 3

Who Created the GLM

Answer 3

Jacob Cohen

Question 4

How did GLM come to be?

Answer 4

-Multiple Regression had been dealing with continuous variables

-ANOVA had been dealing with categorical

Cohen created the GLM to have a regression that can handle categorical variables

He used dummy coding (0, 1) to include group info into equations

Significance: You can now incorporate any predictor in the model

Question 5

Relationship between T-test, One-Way ANOVA, and Factorial

Answer 5

-T-tests are a special case of One-Way ANOVA

-One-way ANOVA is a special case of Factorial ANOVA

-ANOVA is a special case of Regression/GLM

Question 6

Difference between GLM and Generalized Linear Modeling

Answer 6

-GLM is used for normal number outcomes like test scores or IQ

-Generalized Linear/Multilevel Modeling is used for outcomes like Yes/No or Guilty/Not Guilty