ANCOVA Analysis
A recent study of driving behavior suggests that self-reported measures of high driving skills and low ratings of safety skills create a dangerous combination (Note: those who rate themselves as highly skilled drivers are probably overly confident). Drivers were classified as high or low in self-rate skill based on responses to a driver-skill inventory, then classified as high or low in safety skill based on responses to a driver-aggression scale. An overall measure of driving risk was obtained by combining several variables such as number of accidents, tickets, tendency to speed, and tendency to pass other cars. Enter the following data into SPSS and use anANCOVA to analyze the results.
Self-Rated Driving Skill
Low
High
Driving Safety
Low
6
5
6
3
6
4
5
6
3
2
3
4
3
5
4
1
High
3
4
4
3
3
2
4
5
8
9
9
7
8
9
7
9
1. State the independent variables (factors) and dependent variable in this study. 2. What are the mean driving risk scores for each group? 3. State the null and alternative hypotheses for the ANCOVA. 4. State the numerical results for the ANCOVA (include F, df, p, and 2) and state the statistical decision. 5. Write a paragraph as in a journal article that summarizes and interprets the results of the ANCOVA for this study on driving risk.
MANOVA Analysis
Researchers are interested in the effects of cognitive behavior therapy and behavior therapy on Obsessive Compulsive Disorder. We could compare a group of OCD sufferers after CBT and after behavior therapy with a group of OCD sufferers who are still awaiting treatment (a no treatment condition).
As OCD has both cognitive and behavioral elements, it is not enough to look only at therapy’s influence on behavioral outcomes. Therefore, our study used two dependent measures: the occurrence of obsession-related behaviors (Actions) and the occurrence of obsession-related cognitions (Thoughts).
The dependent variables were measured on a single day and so represent the number of obsession-related behaviors/thoughts in a normal day. Use data set OCD.sav.
1. What is/are our research question(s)?
2. Prescreen the data. a. Does our data meet all the assumptions for a MANOVA?
b. Does the data need to be corrected for outliers and normality? (Copy and paste in to this section the information that helped you reach your conclusion).
c. Does the data demonstrate homogeneity of variance-covariance? (Copy and paste in to this section the information that helped you reach your conclusion).
3. “Run” the MANOVA analysis (i.e., use the printout of SPSS output). a. Which MANOVA test statistic should you use? Why?
4. Using the appropriate group multivariate statistics and ANOVA/post hoc tests, report your results.
