We have data from a field experiment on a frequently purchased brand.We have available 76 weeks of data on a control panel of families and a matched experimental panel from the same town.Both panels are exposed to the same base level of advertising for the first 52 weeks.For the last 24 weeks, the control panel continues to be exposed to the base level while the experimental panel is exposed to twice the level of advertising as the control panel.The dependent variable is weekly sales of the brand undergoing the experiment.This is obtained by aggregating over all families who purchase the brand that week.The independent variables are own price and the chief competitive brand's price.A variable DUM was created which takes the value 0 for the first 52 weeks and 1 for the next 24 weeks.The variable DUM can be considered a PRE/POST variable, i.e., a variable that measures change in the last 24 weeks relative to the first 52 weeks.

The variables are:

Week

SalesEXP (unit sales in Experimental group)

SalesCNTL (unit sales in control group)

BPriceEXP (brand's own price in Experimental group)

BpriceCNTL (brand's own price in control group)

CPriceEXP (competitive brand's price in experimental group)

CPriceCNTL (competitive brand's price in control group)

DUM (defined above).

To verify the accuracy of the data you have, you may compare the means of each variable in the data with the numbers below:

VariableMean

SalesEXP7089.7

SalesCNTL6687.7

BPriceEXP1.213

BPriceCNTL1.218

CPriceEXP1.176

CPriceCNTL1.144

DUM0.316

1. Run a regression of SalesEXP against BPriceEXP and CPriceEXP.Based on the output below, please write out the estimated regression equation. (2 points)

1. Is there evidence to conclude that the model for the experimental group is significant as a whole? Using the output from the regression in part a, set up the appropriate hypotheses and conduct the test.Use 95% confidence. (2 point)

1. Are the signs of the coefficients from the regression in part a intuitive? (2 point)