You have collected data on 4 continuous variables (diastolic BP, age, weight, and average number of minute of exercise per week)you think may be related to total cholesterol levels. Use the attached file cholesterol.xls to perform research on this question.
The regression model
a.) Use Excel to perform a multiple regression using all the independent variables. (There is no answer to give for this step.)
b.) Write the regression model based on your results (include both your significant and non-significant betas in this model).
c.) Report your intercept. Is the intercept meaningful? Why or why not?
d.) Report your betas (slopes). Which of these betas is significant?
e.) Give interpretations for each of your significant betasin terms of the original research question.
Be specific and include appropriate units in your interpretations!
As an example, if your independent variable was weight and your beta for weight was 5, you should say, "For every 1 pound increase in weight, total cholesterol increases by 5 mg/dL, controlling for diastolic BP, age, and average minutes of exercise per week." Don't just say "For every unit increase in weight...".
You have reason to suspect that there may be interaction between the diastolic BP and Minutes of Exercise. In your Excel file, construct an interaction term for these two variables and rerun the multiple regressionwith the interaction term included.
f.) Report the beta for interaction. Isthere significant interaction between A and B?
g.) Should you leave the interaction term in the model? Why or why not?