Identify and construct a business or economic data set suitable for a multivariable regression analysis. Include measured quantities (cardinal numerical variables) that you believe should be related to each other.For a meaningful regression you should have at least 30 observations for both your dependent and independent (at least two) variables.

1. Describe the data.Where do they come from? How are the variables being measured? In what units? Why do think these data will be have a relationship, or be correlated? Do think this is likely to be a "cause and effect" relationship? Do you have a hypothesis about how these data may be related?
2. Do you think your relationships are (in an approximate sense) most likely to be: direct or inverse?Explain why you think this is likely to be the case.
3. What kind of specification do you think will best fit your data? Experiment by the different kinds of OLS regression specifications discussed in class: linear, quadratic, log-linear, and combinations of these.
4. Find the model with the most explanatory power and significant estimates.Describe what your regression output shows regarding the statistical properties of your model.
5. What does this model suggest regarding the relationships between your variables?
6. Do you think your data more or less satisfy OLS regression assumptions?Explain why? Or why not?
7. Based on 6), do you think your analysis is reliable?

Your results in a Word Document along with a spreadsheet of your data and regression output.