Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

arlow Company is an online discount retailer. The CEO wants to better understand

ID: 3303239 • Letter: A

Question

arlow Company is an online discount retailer. The CEO wants to better understand their distribution costs, and has asked his accounting team to look into it. Discussion with the workers in the distribution department have determined that distribution expense are operationally associated with:

Number of orders

The number of parcels in an order

Number of large items shipped

Data for the past 24 months as compiled by the accounting department:

Using the data regression methodology from Chapter 3, determine:

What is the best dependent variable to use as a predictor? (Run 3 single variable regressions)

What variables (if any) have little to no effect on distribution cost? (Analyze single variable regression output)

What is the best single variable equation?

What is the best multiple variable equation? (Run Multiple regression analyses)

Using your equation from 3 and 4, If Harlow expects to have 95,000 orders, with 150,000 parcels, and 9,500 large items, what is distribution cost?

What other factors do you think might play a part in distribution cost that are not outlined?

Month Distribution Cost Number of Orders Number of parcels in an order Number of Large items 1 $        45,000          11,200                     38,000                1,120 2 $        58,000          14,000                     45,000                1,400 3 $       155,000          40,000                     55,000                1,000 4 $       450,000         110,000                     50,000                  850 5 $        90,000          20,000                     38,600                4,000 6 $       126,000          33,100                     68,000                5,500 7 $        90,600          21,000                     50,000                1,800 8 $        54,000          12,800                     30,000                2,000 9 $       175,000          43,860                     85,000                1,500 10 $       287,000          50,000                     75,000                2,500 11 $       350,000          92,700                     85,000                5,000 12 $       425,000          85,000                     60,000               14,000 13 $       110,000          28,220                     35,000                1,500 14 $        95,000          21,200                     50,000                1,100 15 $       160,000          38,560                     70,000                1,500 16 $        85,000          19,630                     35,000                1,340 17 $       135,000          35,800                     40,000                3,000 18 $        75,000          18,900                     25,000                2,000 19 $       125,000          33,070                     45,000                1,900 20 $       175,000          43,420                     70,000                1,430 21 $       150,000          35,720                     66,800                1,750 22 $       138,000          35,300                     62,500                2,200 23 $       260,000          68,000                     85,000                9,000 24 $       325,000          87,750                     63,000               15,250

Explanation / Answer

1) Best would be number of orders as the p-value is low and R^2 is high which means higher proportion of variation in dependent variable is explained by indepedent variable.

2) Looking at the single variable equation, we can see that No of large items has the lowest r squared amongst all the single variable equations. Also, it has the highest p-value in the multiple regression equation. Hence, that variable has the least effect amongst the three on distribution cost.

3) Best single variable equation is No of orders because that has the lowest p-value and also highest r squared.

4) Best multiple variable equation amongst the three above would be No of orders and No of parcels in an order because it has the highest r squared amongst all the three.

5)

equation is -4400.7+3.74215(No of orders)+0.3348(No of parcels in an order)-0.25553(No of large items)

-4400.7+3.74215(95000)+0.3348(150000)-0.25553(9500)=-4400.7+355504.25+50200-2427.535=398876.015

6) Other factors that might affect the distribution cost could be distance to destination and distribution channel and also value of goods (higher value goods might have to be shipped through a premium distributor)

P.S:- I am not able to post the excel output (As answer can only be 65000 characters long ) for all of them but posting for all the three variables (multiple regression) and individual variables below. Rest should be combination of two independent variables (so other than this would be 3). The same can be calculated from data analysis tab in excel.

a) All three as independent variables:-

b) No of orders as independent variable:-

c) No of parcels in an order as independent variable:-

Regression Statistics Multiple R 0.974808 R Square 0.95025 Adjusted R Square 0.942788 Standard Error 27692.13 Observations 24 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0% Intercept -4400.7 13417.83 -0.32797 0.74634 -32389.8 23588.41 -32389.8 23588.41 No of orders 3.74215 0.621881 6.017474 0.00 2.44493 5.03937 2.44493 5.03937 No of parcels in an order 0.3348 0.488282 0.68567 0.500793 -0.68374 1.353338 -0.68374 1.353338 No of large items -0.25553 1.796565 -0.14223 0.888319 -4.0031 3.492038 -4.0031 3.492038

b) No of orders as independent variable:-

SUMMARY OUTPUT Regression Statistics Multiple R 0.974199 R Square 0.949064 Adjusted R Square 0.946749 Standard Error 26716.4 Observations 24 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0% Intercept 1434.809 10053.88 0.142712 0.887817 -19415.7 22285.28 -19415.7 22285.28 No of orders 4.107327 0.202868 20.24633 0.00 3.686605 4.528049 3.686605 4.528049

c) No of parcels in an order as independent variable:-

SUMMARY OUTPUT Regression Statistics Multiple R 0.925496 R Square 0.856543 Adjusted R Square 0.850022 Standard Error 44835.88 Observations 24 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0% Intercept -32845.7 20114.41 -1.63294 0.12 -74560.4 8869.08 -74560.4 8869.08 No of parcels in an order 3.128386 0.272957 11.46108 0.00 2.562307 3.694466 2.562307 3.694466 d) No of large items as independent variable:-