EEB225H1 Lecture Notes - Lecture 17: Analysis Of Covariance, Statistical Process Control, Biostatistics
Document Summary
Cumulative, but focus on what has not been tested yet. Allows one to exert statistical control over results. Different from experimental control, which refers to procedural modification. Y = a + x where a is categorical and x is continuous. An x variable that varies with y. May or may not vary with another x variable, which also makes it a confounder. Wish to control effect of covariate on y variable in order to better understand independent effect of x variable of interest. Like 1 way except covariate added to model. Test effect of groups after accounting for variance explained by covariate. 2: regression to see if covariate explains any of the variation in the data. 3: make sure slopes of regression for each level of categorical variable are the same.