Confidence Interval for the Expected value.
You have been provided with the following regression output. The dependent variable is unit sales of a product, and the independent variable is the dollar value of newspaper advertising for that product. The unit of analysis is weeks; that is, each of the observations in the underlying sample data represents a separate week of sales and newspaper advertising.
Regression Analysis: Sales versus Newspaper
the regression equation is
Sales = 6514 + 0.0386 Newspaper
Predictors

Coeff

SECoeff

T

P

Constant

6514

1138

5.72

0.000

Newspaper

0.038625

0.005494

7.03

0.00

S = 5630.70 RSq = 24.3% RSq(adj) = 23.8%
Analysis of Variance
Source

D F

S S

M S

F

P

Regression

1

1567134272

1567134272

49.43

0.000

Residual Error

154

4882532724

31704758



Total

155

6449666996




1. How much would sales increase if Newspaper spending increases by 1?, 1000?
2. How much variability would you have if you predict sales with this model?
3. What would be your best point estimate for average Sales in weeks for which Newspaper advertising were $1 million? What would be an appropriate interval estimate for this? (Use 95% limits.)