regressiogn mastery problem (1)

1
Fin 615 Regression Mastery Problem – Session 5 A senior financial analyst with Ace Gadgets (AG) is attempting to get a better grasp on sales forecasting for AG’s new franchises. She has obtained various details for 27 existing franchises including (see associated spreadsheet): Sales (Annual) Square feet of the store Inventory Advertising $ (annual) Number of competing stores in the district From her recollection of her undergraduate course in statistics, she thought of regression analysis as a possibility in modeling new franchise sales. She has enlisted your help in this modeling task and has provided you with this list of questions. 1. What is the correlation between the above variables and sales? 2. Which variable appears to have the strongest relationship with sales? Why do you suggest this variable? 3. Create a scatterplot between the variable that you selected in requirement 2 and sales. Properly label your chart. 4. Add a trend line to the requirement 3 chart along with the regression equation and R 2 . 5. Interpret (in layman’s language) what the equation means and what the R 2 means. Remember that the senior analyst (senior ≈ old) hasn’t had a course in statistics in several years and needs an interpretation that is understandable. Be sure to include all elements of the equation. 6. Using the analysis toolpak add-in, run regression analysis using the variable that you selected in requirement 2. 7. Using the output from requirement 6, is this variable statistically significant in predicting sales? What specifically on the output allows you to reach this conclusion 1 ? 8. Which variables from the above list are useful in predicting sales? Why? 9. Using an appropriate Excel function, if a new franchise decided to carry $300,000 in inventory, what can be the expected annual sales for this franchise? Are you 100% confident in your answer? Why or why not? 1 Note: In the Mayes text on page 157, Mayes states “a general rule of thumb . . . for large samples, a t-statistic greater than about 2.00 is significant at the 95% confidence level.” I personally do not look at the t-statistic but rather at the p-value – this is much more useful than an absolute t-statistic value. Copyrighted 2012

Upload: michaelkattan

Post on 25-Nov-2015

46 views

Category:

Documents


0 download

DESCRIPTION

g

TRANSCRIPT

  • Fin 615

    Regression Mastery Problem Session 5

    A senior financial analyst with Ace Gadgets (AG) is attempting to get a better grasp on sales forecasting

    for AGs new franchises. She has obtained various details for 27 existing franchises including (see

    associated spreadsheet):

    Sales (Annual)

    Square feet of the store

    Inventory

    Advertising $ (annual)

    Number of competing stores in the district

    From her recollection of her undergraduate course in statistics, she thought of regression analysis as a

    possibility in modeling new franchise sales. She has enlisted your help in this modeling task and has

    provided you with this list of questions.

    1. What is the correlation between the above variables and sales?

    2. Which variable appears to have the strongest relationship with sales? Why do you suggest this

    variable?

    3. Create a scatterplot between the variable that you selected in requirement 2 and sales.

    Properly label your chart.

    4. Add a trend line to the requirement 3 chart along with the regression equation and R2.

    5. Interpret (in laymans language) what the equation means and what the R2 means. Remember

    that the senior analyst (senior old) hasnt had a course in statistics in several years and needs

    an interpretation that is understandable. Be sure to include all elements of the equation.

    6. Using the analysis toolpak add-in, run regression analysis using the variable that you selected in

    requirement 2.

    7. Using the output from requirement 6, is this variable statistically significant in predicting sales?

    What specifically on the output allows you to reach this conclusion1?

    8. Which variables from the above list are useful in predicting sales? Why?

    9. Using an appropriate Excel function, if a new franchise decided to carry $300,000 in inventory,

    what can be the expected annual sales for this franchise? Are you 100% confident in your

    answer? Why or why not?

    1 Note: In the Mayes text on page 157, Mayes states a general rule of thumb . . . for large samples, a t-statistic

    greater than about 2.00 is significant at the 95% confidence level. I personally do not look at the t-statistic but rather at the p-value this is much more useful than an absolute t-statistic value.

    Copyrighted 2012