APST 207 Lecture Notes - Lecture 13: Random Assignment, Analysis Of Variance, Null Hypothesis
Document Summary
One-way anova: using only 1 independent variable (with at least three classes, or levels) and 1 independent variable. Groups need to be independent of one another. Randomly assign a nutrition, exercise, control group to measure healthiness. Repeated measures anova: each case is measured multiple times and their means are compared. Uses multiple independent and dependent variables and each case has a score on each variable. Groups are not independent of one another. Normally distributed y (dependent variable) for every x (independent variable) Population variances for y at every level/group/ value of x are equal. You can get away with violating the first assumption a little. A random assignment is important, but naturally occurring groups can be used. We need normal distributions and equal variance in each cell. It gives you significant results that are not really significant. Anovas were born out of a necessity to test experiments.
