Statistical Sciences 2035 Quiz: Split-Plot
16 Oct 2018
School
Department
Professor
Document Summary
There are 2 typical varieties of the split-plot design: ws design in which we also include a characteristic of the participant (e. g. gender) Example: bs factor: teaching method, ws factor: each follow-up statistics test. This design was used in the example from the lecture (brug"s study on the effects of nutrition education) We are only looking at 4 boxes in blue square 2x2 split-plot design. Yes, if frequency distributions are equal i. e. if equal amount of people in each group are measured in post1 and post2: Without drop-out (attrition), split-plot designs should always be orthogonal. Split plot is robust to violations of normality if central limit theorem (clt) applies: we need more than 20 participants. True here, so we can carry on. They are 3 standard deviations away from the mean check descriptives table & histograms. This rule assumes normality so it"s not valid when we have a violation of normality (like here)
