BIO361H5 Lecture 3: Lec 3 - Sept 13
Document Summary
Assuming equal variances (the treatment groups should have comparable variances) Conclude there"s a significant difference b/t eye-span and diet. How to figure out variances are equal: f-test, use box plots (if boxes are same size), using s. d of both groups and divide them, in r - var. test(eyespan~diet) Use it with many number of levels you"re comparing between groups (ex. if you have 3 diets instead of 2) Looks at variation between the groups not within the groups. Reason we don"t use multiple tests reduce chance of type 1 error (might reject null hyp when there is no difference) In case of more than 2 levels, don"t do multiple t-tests. In anova, it calculates mean of each group and an overall mean. When already attached data set: mean(eyespan) instead of mean(c2. stalkeyed. diet==) Linear model: shows how much change is occurring between the means.
