You are in charge of marketing for a fast casual restaurant chain.Fortunately, your restaurants have been able to remain in operation during the pandemic.All of your locations have drive-through windows and you already offered local delivery in a five mile radius. You have been tasked with determining the impact of COVID19 lockdowns on unit sales for your 350 locations.The dataset "2020 Sales" has information on Q2 sales for each unit.The variables are:

Competition Index:An index that measures how much competition surrounds each unit.The scale is such that 50 is average, anything less than 50 means less than average competition, and anything above 50 means more than average competition.

Population:The number of people that live within a 5 mile radius of the unit.

Affluence:An index that measures how much affluence surrounds each unit.The scale is such that 50 is average, anything less than 50 means less than average affluence, and anything above 50 means more than average affluence.

PreviousSales:Unit sales in Q2 of 2019.

Lockdown:Whether there was a complete lockdown or not (1=complete lockdown, 0=partial lockdown).

BrandRecog:A number which represents the percent of people in the area of the unit that list your restaurant in unaided brand awareness studies (e.g. 42 would mean that 42% of people list your restaurant in unaided brand awareness studies).

CurrentSales: Unit sales in Q2 of 2020.

Please use this information to answer the following questions.

Please examine the distributions and correlations to select the best answer for the following questions (Ignore UnitNumber as its is just a unit id):

Choose from the list below:

39.54; Previous Sales; False, Lockdown; Population; 39.47; True

1.Which variable is not normally distributed?

2.What is the Median level of unaided brand recognition?

3.Based on the correlations alone, it appears as though units with better brand recognition have higher current sales.

4.We can easily see that the pandemic has negatively impacted sales as they are down since the same quarter last year.

5.Which variable has the highest correlation with Q2 2020 sales?

Now build a multiple linear regression using Current Sales as the Y and Competion Index, Population, Affluence, Previous Sales, Lockdown, and Brand Recognition as predictors and select the best answer for the following questions:

Choose from the list below:

Affluence; True; Population; 97.42%; False; 2204.515; Previous Sales; 97.47%

1.Which variable has the highest p value?

2.The model is problematic because it has non-constant variance?

3.How much of the variance in Current Sales is explained by the model?

4.The overall model is statically significant at the .01 level of significance?

5.The intercept is meaningful in the context of this problem?

Manually make any adjustments to the model you deem necessary (do not use stepwise at this point) in order to build the most parsimonious model.Please use your results to answer the following questions:

Choose True/False

1.The resulting model meets the assumption of constant variance.

2.Brand Recognition and Affluence are significant predictors of Current Sales.

3.The final model explains more than 97% of the variation in Current Sales.

4.Units in areas with full lockdowns (as opposed to partial lockdowns) lost over $50,000 per unit in Q2 sales.

5.Previous Sales is a statistically significant predictor of current sales.

Instead of simplifying the model manually, run a forward stepwise regression with minimum BIC as the stopping rule and direction as forward.Include Current Sales as the outcome and all other variables (except Unit Number) as predictors. Based on this model, please answer the following questions:

Choose from the list below:

Lockdown, Competition Index; Competition Index, Populations, Affluence; Previous Sales, Lockdown; True; False; Competition Index, Previous Sales, Lockdown

1.The resulting model meets the assumptions of linear regression.

2.Which predictors are contained in the resulting model?

3.All of the predictors in the model are statistically significant.

4.The resulting model explains more than 95% of the variation in Current Sales.

Question text

Run a second stepwise regression using Minimum AIC stopping rule, forward direction, Current Sales as the outcome and all other variables (except Unit Number) as predictors. Use the results to answer the following questions. (For the purposes of these questions, this model will be referred to as Model 2 and the previous model run using Minimum BIC and forward direction will be referred to as Model 1).

Choose 1:

Not Enough Information; True; False

1.Model 2 has the same number of variables as Model 1.

2.The r square of Model 2 is lower than Model 1.

3.All variables in Model 2 are statistically significant at a .05 level.

Based on your analysis so far, please answer the following questions as True, False, or Not Enough Information

Choose 1:

False; True; Not Enough Information

1.Full lockdowns, in comparison to partial lockdowns, reduced overall unit sales (across all 350 locations) by more than $10mm in Q2 of this year.

2.No matter how we build the model, we end up with the exact same set of variables.

3.The pandemic has hurt overall sales.