William H. Knapp III

You will not be able to submit your work for credit, because you are not logged in. Log in!

This homework was due on Friday, January 4 at 06:00 a.m. Turkish time. Late submissions receive half credit.

By checking the box below, you certify that the answers you will submit here represent your own work.

1. What type of tests are tests like ANOVA?
Non-parametric
Parametric
Impossible to tell

2. What type of tests are tests like Kruskal-Wallis and Friedman tests?
Non-parametric
Parametric
Impossible to tell

3. Imagine you wanted to test the effects of different types of therapy on well-being. To do this you put out a newspaper add offering free therapy and then gave the respondants 3 months of free therapy. For each respondant you randomly determined whether they would receive psychoanalysis, client-centered therapy, or cognitive behavioral therapy. At the end of the 3 months you measured participants well-being on some scale that is not normally distributed. Because you only have a small number of observations for each group, what test should you use?
Dependent-samples Wilcoxon
Friedman
Independent-samples Wilcoxon
Kruskal-Wallis
One-sample Wilcoxon
Impossible to tell

4. In the last situation you had 3 types of therapy. If you had 30 participants, how many degrees of freedom would you use in determining the significance of the previous test?

5. The previous design isn't as powerful as it could have been, imagine you took well-being scores before and after each respondant received their 3 months of therapy. You then created a change of well-being score by subtracting the before scores from the after scores. Since you're not sure that the change of well-being scores are normally distributed, what test should you use?
Dependent-samples Wilcoxon
Friedman
Independent-samples Wilcoxon
Kruskal-Wallis
One-sample Wilcoxon
Impossible to tell

6. Imagine you and five friends wanted to test different study habits. Each of you and your five friends tried each one of 4 different study techniques before taking tests throughout the semester. At the end of the semester, you wanted to see if one study technique was more effective than the other. Because each of you were taking tests in different classes and have rather small sample sizes what test should you use?
Dependent-samples Wilcoxon
Friedman
Independent-samples Wilcoxon
Kruskal-Wallis
One-sample Wilcoxon
Impossible to tell

7. In the last situation, imagine you and your five friends (i.e. 6 students total) took tests after studying with 4 different methods. How many degrees of freedom would you use in determining the significance of the previous test?

8. ANOVAs, Friedman, and Kruskal-Wallis tests can all be expressed as the ratio of weighted sums of squares. For ANOVA, what sum of squares appears in the numerator?
Effect Sum of Squares
Error Sum of Squares
Total Sum of Squares
Impossible to tell

9. ANOVAs, Friedman, and Kruskal-Wallis tests can all be expressed as the ratio of weighted sums of squares. For ANOVA, what sum of squares appears in the denominator?
Effect Sum of Squares
Error Sum of Squares
Total Sum of Squares
Impossible to tell

10. ANOVAs, Friedman, and Kruskal-Wallis tests can all be expressed as the ratio of weighted sums of squares. For Friedman tests, what sum of squares appears in the numerator?
Effect Sum of Squares
Error Sum of Squares
Total Sum of Squares
Impossible to tell

11. ANOVAs, Friedman, and Kruskal-Wallis tests can all be expressed as the ratio of weighted sums of squares. For Friedman tests, what sum of squares appears in the denominator?
Effect Sum of Squares
Error Sum of Squares
Total Sum of Squares
Impossible to tell

12. ANOVAs, Friedman, and Kruskal-Wallis tests can all be expressed as the ratio of weighted sums of squares. For Kruskal-Wallis tests, what sum of squares appears in the numerator?
Effect Sum of Squares
Error Sum of Squares
Total Sum of Squares
Impossible to tell

13. ANOVAs, Friedman, and Kruskal-Wallis tests can all be expressed as the ratio of weighted sums of squares. For Kruskal-Wallis tests, what sum of squares appears in the denominator?
Effect Sum of Squares
Error Sum of Squares
Total Sum of Squares
Impossible to tell

14. In a Friedman test, what do you multiply the ratio of the weighted sums of squares by?
Number of groups minus one.
Number of observations minus one.
Number of observations per groups minus one.
Number of observations per group times the number of groups minus one.

15. In a Kruskal-Wallis test, what do you multiply the ratio of the weighted sums of squares by?
Number of groups minus one.
Number of observations minus one.
Number of observations per groups minus one.
Number of observations per group times the number of groups minus one.

16. In a Kruskal-Wallis test, how are the ranks determined?
All of the data are ranked together.
The data are ranked separately for each 'group'.
The data are ranked separately for each 'participant'.
The data are ranked separately for each 'participant' in each 'group'.

17. In a Friedman test, how are the ranks determined?
All of the data are ranked together.
The data are ranked separately for each 'group'.
The data are ranked separately for each 'participant'.
The data are ranked separately for each 'participant' in each 'group'.

18. What is the null hypothesis for the Friedman and Kruskal-Wallis tests?
The median of the scores is 0.
The median of the scores is equal to some value.
The medians/means of the groups are all the same.
The medians/means of the groups are all different.

19. In class, we analyzed these data to look and see if different 'sad' movies were rated differently. The data are in the form of a data frame that includes ratings, judges, and movies (use str to see what the variables are named if you need to). Let's look at the same data and see if the different judges rate movies differently (i.e. do some judges rate movies as sadder than others). In class we treated the movies as our groups and the judges as our participants. Here we want to treat the judges as our groups and the movies as our participants. What is the probability that of observing a test result as extreme as you did?
HINT: In class, when we wanted to see if there was a difference in the movies we used the following model: ratings~movie. If you adjust the model and perform the right test, you should find a Chi-square of 8.2.

20. Knowing that each judge judged the same movies we can do a more powerful test and control for some of the variance in ratings. What is the probability of observing a test result as extreme as you did?
HINT: If you did the right test on the right model, you should find a Chi-square of 10.9333.