1. A total of one hundred and twenty newly hired employees are equally divided by random assignment to one or another of five training conditions that differ in terms of the number of hours of training they receive per day (one, two, three, four, or five). Performance is measured on a 100-point scale for each trainee after three months of training. The means for each condition are shown in the table below.

One Two Three Four Five

Mean 50 60 65 68 70

Given that SStotal is 80,000 complete the one-way ANOVA, and test your results at the .05 level. Show the results.

a. Suppose that you decided to replicate this experiment using n = 36 subjects. What would power be?

b. Use Fishers and Tukeys test to determine which groups are significantly different from each other and what can you conclude from the results? Is Fisher's justified here and why.

2. Calculate the linear contrast to test whether or not 3,4 and 5 hours of training combined differs from 1 and 2 hours of training combined from question 1 above.

3. Calculate the LINEAR TREND contrast to test whether hours of training shows a trend increase using the coefficients in the statistics tables.

4. A psychologist believes that people who exercise regularly have higher self-esteem than those who do not. To prove her hypothesis, she selects eight participants and randomly assigns each to a different number of hours of exercise per week. After six months of following the assigned exercise regimen, she measures each participant's self-esteem using an established questionnaire. The data collected on the eight participants appear in the table below.

Subject #

2

4

5

7

6

8

9

3

Exercise Hours Per Week

2

4

5

7

6

8

9

3

Self-Esteem

28

25

26

30

37

33

36

29

a.Calculate the Pearson correlation coefficient for the data above.

b.If someone exercised 15 hours per week, what would their self-esteem score be?

c.How wrong would you be?