EEB225H1 Lecture Notes - Lecture 16: Analysis Of Covariance, Factorial Experiment, Biostatistics
18 Dec 2019
School
Department
Course
Professor
Document Summary
E. g. 2 factor anova, multiple regression, ancova. A model is a mathematical representation of the relationship between a response variable y and one or more explanatory variables. To test for more than 1 explanatory factor and interactions between categorical explanatory variables (factorial design) and continuous explanatory variables (multiple linear regression) To improve detection of treatment effects (blocking) To control for non-independent of subjects (nesting) 2 or more categorical explanatory variables investigated. Y(numeric) = constant + a(categorical) + b(categorical) 2 way factorial = 2 predictor variables. But sometimes 2 way can be described in terms of the number of levels . E. g. sex (2 levels) by age (3 levels), a 2 by 3 way factorial ancova. Interactions: score = age + sex + age*sex. Partition variance in different ways to test 3 alternate hypotheses. Like a 1 way, except instead of comparing 1 set of group means, can evaluate data grouped in different ways.
