Final material

Question 1

When to use…
1. Only a post HOC test
2. Both post HOC test and simple main effects
3. Just simple main effects.

Answer 1

1. Only post-HOC: If the interaction is not significant but at least one of the main effects is
2. Post-HOC and simple main effects: If the interaction is significant but it is dominated by the main effect
3. Only simple main effects: If the significant interaction dominates the main effect

Question 2

What are 4 ways you could select which simple main effects to examine if it’s unclear (according to Keppel and Wickens)

Answer 2

1. Choose the factor with the greatest number of levels
2. Choose a quantitative factor
3. Choose the factor with the greatest SS for the main effect: (maximizes the amount of variability that the simple main effects)
4. Choose an experimentally manipulated factor

Question 3

What does the notation N_mean signify for simple main effects

Answer 3

It signifies the number of observations that contributed to each mean

Question 4

After simple main effects, when do you have to compute a simple comparison?

Answer 4

If you have a simple main effect that has more than one degree freedom in the numerator,(meaning a comparison of more than two groups), and that simple main effect is significant, you have to follow it up by what’s called a simple comparison.

Question 5

The greater the violation of sphericity, the (more/less) conservative you have to make your F ratio, and the (smaller/bigger) the value of epsilon needed to adjust degrees of freedom.

Answer 5

The greater the violation of sphericity, the more conservative you have to make your F ratio, and the smaller the value of epsilon needed to adjust degrees of freedom.

Question 6

Assign each trait to either Scheffé’s test or Tukey’s HSD:
1. Less conservative
2. Good for unbalanced designs
3. Better control for Type 1 error but less power
4. Uses the Q statistic
5. Less sensitive to violations of assumptions

Answer 6

Scheffé:
Better control for Type 1 error but less power, good for unbalanced designs, less sensitive to violations of assumptions

Tukey:
Less conservative, more power, uses Q statistic

Question 7

In a within subjects design, what research question are we mostly interested in (and which one are we NOT)

Answer 7

Interested in: Are there differences in the mean scores of the DV across groups/conditions?
NOT interested in knowing wether there’s differences in the mean score of DV across subjects