# Question

Do the following analysis using the hypothetical data in Table 13.7. In this imaginary experiment, participants were randomly assigned to receive either no caffeine (1) or 150 mg of caffeine (2); and to a no exercise condition (1) or half an hour of exercise on a treadmill (2). The dependent variable was heart rate in beats per minute.

a. Graph the distribution of HR scores (as a histogram). Is this distribution reasonably normal? Are there any outliers?

b. Compute the row, column, and grand means by hand. Set up a table that shows the mean and n for each group in this design.

c. Calculate the α,β, and αβ effect estimates and report these in a table (similar to Table 13.5).

d. Comment on the pattern of effects in the table you reported in part 2c. Do you see any evidence of possible main effects and/or interactions?

e. Set up an Excel or SPSS worksheet that breaks each score down into components, as in the spreadsheet shown in Table 13.4.

f. Sum the squared effects in the spreadsheet you just created to find your estimates for SSA (main effects of caffeine), SSB (main effects of exercise), and SSA x B (interaction between caffeine and exercise).

g. Run a factorial ANOVA using the SPSS GLM procedure. Verify that the values of the SS terms in the SPSS GLM printout agree with the SS values you obtained from your spreadsheet. Make sure that you request cell means, a test of homogeneity of variance, and a plot of cell means (as in the example in this chapter.)

h. What null hypothesis is tested by the Levene statistic? What does this test tell you about possible violations of an assumption for ANOVA?

i. Write up a Results section. What conclusions would you reach about possible effects of caffeine and exercise on heart rate? Is there any indication of an interaction?

a. Graph the distribution of HR scores (as a histogram). Is this distribution reasonably normal? Are there any outliers?

b. Compute the row, column, and grand means by hand. Set up a table that shows the mean and n for each group in this design.

c. Calculate the α,β, and αβ effect estimates and report these in a table (similar to Table 13.5).

d. Comment on the pattern of effects in the table you reported in part 2c. Do you see any evidence of possible main effects and/or interactions?

e. Set up an Excel or SPSS worksheet that breaks each score down into components, as in the spreadsheet shown in Table 13.4.

f. Sum the squared effects in the spreadsheet you just created to find your estimates for SSA (main effects of caffeine), SSB (main effects of exercise), and SSA x B (interaction between caffeine and exercise).

g. Run a factorial ANOVA using the SPSS GLM procedure. Verify that the values of the SS terms in the SPSS GLM printout agree with the SS values you obtained from your spreadsheet. Make sure that you request cell means, a test of homogeneity of variance, and a plot of cell means (as in the example in this chapter.)

h. What null hypothesis is tested by the Levene statistic? What does this test tell you about possible violations of an assumption for ANOVA?

i. Write up a Results section. What conclusions would you reach about possible effects of caffeine and exercise on heart rate? Is there any indication of an interaction?

## Answer to relevant Questions

Write out an equation to represent a 2 x 2 factorial ANOVA (make up your own variable names); include an interaction term in the equation. Describe a hypothetical study for which multiple regression with more than two predictor variables would be an appropriate analysis. Your description should include one dependent variable and three or more predictors. a. For ...What types of research situations often make use of multiple regression analysis with more than two predictors? What kinds of preliminary data analysis are helpful in assessing whether moderation might be present? (Additional information about this is provided in Chapter 10). When adjusted group means are compared with unadjusted group means, which of the following can occur? (Answer yes or no to each.) • The adjusted group means can be closer together than the unadjusted group means. • The ...Post your question

0