saptamana treispe

Question 1

What is an advantage of using related ANOVA?

Answer 1

Variation due to individual differences can be separated from pure error variation

We have less errors due to individual differences

Question 2

What are the assumptions?

Answer 2

One additional: homogeneity of covariance
Correlations between scores in different conditions are similar: participants with highest score in one condition will have the highest score in another condition

Question 3

What does heterogeneity of covariance increase?

Answer 3

Type 1 error rate

Question 4

What is Mauchly’s sphericity test?

Answer 4

If Mauchly is significant then the assumption has been violated
We don’t want Mauchly to be significant

If assumption is violated = Greenhouse-Geisser test reduced degrees of freedom associated by F test by factor which depends on amount of hetereogeneity in data

Question 5

How to report data?

Answer 5

If Mauchly is not significant: use first line (sphericity assumed)
If Mauchly is significant: use Greenhouse Geisser (second line)

Question 6

What are the post hoc test of related ANOVA?

Answer 6

ANOVA is an omnibus test

Formula for determining total number of pairwise comparisons
(G X (G-1))/2
Where g is the number of levels of IV

Question 7

What is Bonferroni correction?

Answer 7

Carrying out multiple tests on the same data - increases the chance of type 1 error where we get a false positive - we think we found significant results but they are not real

Bonferroni correction changes significance level to look for

Question 8

What are the methods for Bonferroni correction?

Answer 8

1. Adjust alpha level - adjusting the significance level we are looking for
   Divide the p by the number of comparisons you are doing
2. Adjust observed p values
   Multiply each of the obtained p value and multiply by the total number of comparisons
   The new corrected p values are compared to usual level of 0.05