1. Perform the appropriate multivariate technique to identify underlying dimensions that may exist among the emotion ratings.

2. Create scales for any underlying dimensions.

3. Use these scales as independent variables in a regression model.

4. Interpret the results.

Use the data labeled profit for this case. The data go along with The Research Snapshot on page 587. In addition, management has collected several semantic differential scales from the managers asking them to use emotions to describe the way they feel about their jobs. The emotions include involved exciting fun satisfied happy pleased. The managers want to understand turnover. So, another variable is included that gives the likelihood a manger will quit within 12 months (labeled turnover in data). After running some initial regression models with eight independent variables predicting turnover, management was confused. They complained that there were too many variables to make sense of.