Assessing Relationships
Due 11/27

The final groups of statistical procedures that we will cover deal with situations where you are interested in knowing whether two or more variables are "related"; in other words, is one variable affected by changes in the other variable.  We are often interested in these kind of relationships in marketing.  For example, we may want to know what will happen to sales volume if we raise our price by 10%, or what will happen to customer satisfaction if we shorten the length of happy hour.

The two statistical techniques that we use for these projects are Correlation Analysis and Regression Analysis.  In both procedures, ALL variables of interest must be either INTERVAL or RATIO.  Correlation is used to simply test for the existence of a relationship between two or more variables.  There is no specification of Independent and Dependent variables, so no attempt is made to determine which variable "causes" a change in the other variables.  In Regression, there is a separation of variables into dependents and independents, so we are trying to determine whether changes in one variable (the independent) leads to changes in the other (the dependent).

The following questions will require you to run either correlation analysis or regression analysis in order to find the answers.  Run the appropriate procedure and use the results to support your answers.

1.     In our questionnaire we asked respondents to rate the importance of having classes on campus and of having face-to-face instruction from instructors. It would seem logical that students who rate these things important would also rate faculty, facilities, and class size highly because they would be more likely to actually be in class. What do you think? Are there any relationships among these variables? If so, what is the nature of the relationships?

2.    If we are to improve the overall satisfaction of UAH Business students, we need to understand the things that impact satisfaction the greatest. One of the things that would seem logical is that student satisfaction would increase if we offer the right kinds of courses.  In our questionnaire we asked students to identify the things that would be important to their perfect course.  Of those variables, which ones are the best predictors of overall student satisfaction with the UAH Business school?