BIOL 373 Lecture 15: BIOL 373 LECTURE 15
5 Mar 2018
School
Department
Course
Professor
Document Summary
Two factor anova without replication: only measured one individual for each factor, only one point in each cell, ss cells = ss total, ss error still exists, but cannot be measured, therefore isn"t. F- ratios: 2 way without replication: interaction cannot be tested because ms error is not separate from ms interaction, if (theoretically) interaction isn"t possible, then ms remainder = ms error = ms interaction. Works for all models: if interactions is present, then ms remainder = ms interaction doesnt equal ms error. More anovas based on biological questions with same statistical basis: randomized block design, repeated measures design. Interested in variation within but not between subjects. Advantages of repeated measures: similar to paired sample t test, differences among subjects is removed, may be more powerful than a single factor anova. Disadvantages of repeated measures: treatment order effects, does sequence of treatments matter, carry over effect, early treatments can affect later ones.
