Week 5

Question 1

What are the types of predictors discussed in the lecture?

Answer 1

Categorical (treatment levels A, B and C) and continuous (initial bacillus levels)

Question 2

What are the 3 hypotheses involved in ANCOVA?

Answer 2

Variate, factor, and interaction between them

Question 3

How can you identify a balanced or unbalanced design?

Answer 3

Check for NA’s in the tapply output; unbalanced designs may require adjusted sums of squares

Question 4

What should be checked after running an analysis?

Answer 4

Check assumptions to ensure the validity of the analysis

Question 5

How are GLM and regression equations used in ANCOVA

Answer 5

Write out equations for each line based on summary output to interpret results

Question 6

What is AIC used for in model selection?

Answer 6

Helps decide which variables to include by measuring how well they explain variance

Question 7

When should adjusted sums of squares be used?

Answer 7

Use adjusted sums of squares for unbalanced data without a specific hypothesis order.

Question 8

How can statistical power be increased in a study?

Answer 8

By increasing sample size, including new variables that explain variance, or removing variables with little explanatory power.

Question 9

What is pseudoreplication?

Answer 9

It involves using treatments that are not replicated or replicates that are not statistically independent.