One-Way ANOVAs

Question 1

Why would you use one-way ANOVAs instead of just running a t-test?

Answer 1

• T-Tests give you the same value as a regression but can only be used to compare 2 means
• We’ve got more than 2 groups

Question 2

What is the danger of multiple comparisons?

Answer 2

There is a higher chance of a type 1 error

Question 3

What is a type 1 error?

Answer 3

• Being told something is significant when it is not

- False positive

Question 4

What is a type 2 error?

Answer 4

• Being told something is not significant when it is

- False negative

Question 5

What is a familywise error?

Answer 5

Increased likelihood for a type 1 error

- Can be worked out with 1-(0.95)^k where k is the number of comparisons

Question 6

What does the familywise error rate show us about t-tests?

Answer 6

Conducting many t-tests is a bad idea as you will likely ‘find effects’ that are not really there

Question 7

What is the difference between a one-way ANOVA and a factorial ANOVA?

Answer 7

There is ONE independent variable (can have more than 2 levels)

Question 8

What are the assumptions of a One-Way ANOVA?

Answer 8

• Data is independent
• Normality
• Homogeneity of variance

Question 9

What does a big F ratio suggest?

Answer 9

There is a big proportion of the variance explained by the model

Question 10

What is the MSm? (Mean Squares of the Model)

Answer 10

The variance explained by the model (systematic variance)

Question 11

What is the MSr? (Mean Squares of the Residuals)

Answer 11

Variance left over (noise)

Question 12

What follow up tests can you run if your ANOVA is significant?

Answer 12

• Planned Contrasts

- Post-hoc tests

Question 13

What are planned constrasts?

Answer 13

• Used for testing specific hypotheses

- Make a smaller number of comparisons but can include multiple conditions in each comparison