The data on candy bars per serving from Exercise 16 also has information on whether the candy bar has nuts. The variable Nuts is coded 0 for candy bars without nuts and 1 for those with nuts. Here’s the regression model of Exercise 16 with Nuts added:

a) What is a proper interpretation of the coefficient of Nuts in this regression?
b) Did adding Nuts improve the regression model?

Data From Exercise 16

A student collected nutrition data about candy bars by reading the labels in a supermarket.  Because candy bars have different “serving sizes,” the data are  given as values per serving. Here is a regression predicting  calories  from the sugar (g/serving). (Five sugar-free candy bars have been omitted from this analysis.)

Response variable is: Calories R-squared 46.7% s = 0.3318 Variable Coefficient Intercept 5.74529 Sugar Protein Nuts -2.44376 7.13429 -0.155754