Lecture 11- Mixed Designs

Question 1

What code will extract contrasts

Answer 1

emmeans::contrast()

Question 2

With ‘emmeans::contrast()’, what does ‘eff’ result in

Answer 2

Each category is compared to the average of all categories

This is the default

Question 3

With ‘emmeans::contrast()’, what does ‘pairwise’ result in

Answer 3

Each category is compared to all others

Question 4

With ‘emmeans::contrast()’, what does ‘poly’ result in

Answer 4

Polynomial contrasts (i.e. tests for an ordered trend in category means)

Question 5

With ‘emmeans::contrast()’, what does ‘trt.vs.ctrl’ result in

Answer 5

• Compares each category to a declared reference category (by default the first category)
• Use ‘ref = x’ to make ‘x’ the reference category

Question 6

With ‘emmeans::contrast()’, what does ‘consec’ result in

Answer 6

Compares each level/category (except the first) to the previous

Question 7

Mixed designs combine

Answer 7

At least one predictor measured using repeated measures, and at least one measured using different entities

Question 8

What’s the bare minimum for mixed designs and the linear model

Answer 8

• Modelling variation in ratings overall across participants
• Not modelling variation in the effect of repeated measures predictors across participants