Faculty of Agribusiness, Education and Extension

Department of Agricultural and Applied Economics

AEB 418: Introduction to Econometrics

Assignment 4
Date: 04 November 2024
Due Date: 13 November 2024 (by 12:00)

Instructions: Answer all questions in this assignment. Do not attempt this assignment in the last
minute.
You are each expected to do the assignment independently. Any infringement will lead to loss of
points.
Question 1 (70)

1
The datafile SALES AND PROFITS contains quarterly data on profits and sales for US
Manufacturing for 1965-1970.

a) Using a calculator, compute mean profits for each quarter. (4)

For Q1
10503 + 12245 + 11349 + 12539 + 14151 + 12381 = 73 168
Mean Q1 = 73 168/6
= 12 194.67

For Q2
12092 + 14001 + 12615 + 14849 + 15949 + 13991 = 83 497
Mean Q2 = 83 497/6
= 13 916.17

For Q3
10834 + 12213 + 11014 + 13203 + 14024 + 12174 = 73 462
Mean Q3 = 73,462/6
= 12 243.67

For Q4
12201 + 12820 + 12730 + 14947 + 14315 + 10985 = 77 998
Mean Q4 = 77 998/6
= 12 999.67

b) Estimate the regression:

𝑃𝑟𝑜𝑓𝑖𝑡𝑠𝑡 = 𝛼1 + 𝛼2𝐷2𝑡 + 𝛼3𝐷3𝑡 + 𝛼4𝐷4𝑡 + 𝑢𝑡

where:
D2 = 1 for the second quarter, 0 otherwise
D3 = 1 for the third quarter, 0 otherwise
D4 = 1 for the fourth quarter, 0 otherwise

Report the results. (10)

Profits_t = 12194.67 + 1721.5D₂t + 49D₃t + 805D₄t

c) Using the estimated coefficients, calculate the following:

2
 𝐸(𝑃𝑟𝑜𝑓𝑖𝑡𝑠|𝐷2 = 0, 𝐷3 = 0, 𝐷4 = 0) ii.
𝐸(𝑃𝑟𝑜𝑓𝑖𝑡𝑠|𝐷2 = 1, 𝐷3 = 0, 𝐷4 = 0)
i.

iii. 𝐸(𝑃𝑟𝑜𝑓𝑖𝑡𝑠|𝐷2 = 0, 𝐷3 = 1, 𝐷4 =
0) iv. 𝐸(𝑃𝑟𝑜𝑓𝑖𝑡𝑠|𝐷2 = 0, 𝐷3 =
0, 𝐷4 = 1)
(8)

i.E(Profits|D₂=0, D₃=0, D₄=0) = α₁ = 12194.67

ii.E(Profits|D₂=1, D₃=0, D₄=0) = α₁ + α₂ = 12194.67 + 1721.5 = 13916.17

iii.E(Profits|D₂=0, D₃=1, D₄=0) = α₁ + α₃ = 12194.67 + 49 = 12243.67

iv.E(Profits|D₂=0, D₃=0, D₄=1) = α₁ + α₄ = 12194.67 + 805 = 12999.67

Explain what each estimate means and compare your results with those obtained in
part a) (4)
 The difference of 1,721.5 (α₂) represents how much higher Q2 profits are
compared to Q1
 The small difference of 49 (α₃) shows Q3 profits are only slightly higher than
Q1
 The difference of 805 (α₄) shows Q4 profits are moderately higher than Q1
d) Conduct the following hypothesis tests using any of the methods covered in class.

i. 𝐻0: 𝛼1 = 0
𝐻1: 𝛼1 ≠ 0
ii. 𝐻0: 𝛼2 = 0
𝐻1: 𝛼2 ≠
0 iii. 𝐻0: 𝛼3 =
0 𝐻1: 𝛼3 ≠ 0
iv. 𝐻0: 𝛼4
=0
𝐻1: 𝛼4 ≠ 0
(8)
Interpret the results of each hypothesis test. (8)
i. H₀: α₁ = 0 t = 22.42, p = 0.000 Since p < 0.05, we reject the null hypothesis.
The intercept is statistically significant.
ii. H₀: α₂ = 0 t = 2.24, p = 0.037 Since p < 0.05, we reject the null hypothesis.
Quarter 2 dummy is statistically significant.
iii. H₀: α₃ = 0 t = 0.06, p = 0.950 Since p > 0.05, we fail to reject the null
hypothesis. Quarter 3 dummy is not statistically significant.
iv. H₀: α₄ = 0 t = 1.05, p = 0.308 Since p > 0.05, we fail to reject the null
hypothesis. Quarter 4 dummy is not statistically significant.

3
 e) Estimate the regression:

𝑃𝑟𝑜𝑓𝑖𝑡𝑠𝑡 = 𝛼1 + 𝛼2𝐷2𝑡 + 𝛼3𝐷3𝑡 + 𝛼4𝐷4𝑡 + 𝛽𝑆𝑎𝑙𝑒𝑠𝑡 + 𝑢𝑡

and report the results. (10)

Profits_t = 6688.363 + 1322.892D₂t - 217.8054D₃t + 183.8564D₄t + 0.0382462Sales_t

f) Using your results from part e), derive mean profits for each quarter?
(8)

𝐸(𝑃𝑟𝑜𝑓𝑖𝑡𝑠|𝐷2 = 0,𝐷3 = 0,𝐷4 = 0)= 6688.363

QUATER 1

𝐸(𝑃𝑟𝑜𝑓𝑖𝑡𝑠|𝐷2 = 1,𝐷3 = 0,𝐷4 = 0) =6688.363 +1322.892= 8011.255

QUARTER 2

𝐸(𝑃𝑟𝑜𝑓𝑖𝑡𝑠|𝐷2 = 0,𝐷3 = 1,𝐷4 = 0) =6688.363+217.8054= 6906.168

QUARTER 3

𝐸(𝑃𝑟𝑜𝑓𝑖𝑡𝑠|𝐷2 = 0,𝐷3 = 0,𝐷4 = 1) =6688.363+183.8564= 6872.2194

QUATER 4

g) Suppose the value of sales is 12000. Calculate mean profits for each quarter.
(10)
Quarter 1 (all dummies = 0):
Profits = 6688.363 + 0 + 0 + 0 + (0.0382462 × 12000) = 6688.363 + 458.9544 = 7147.32

Quarter 2
Profits = 6688.363 + 1322.892 + 0 + 0 + (0.0382462 × 12000) = 6688.363 + 1322.892 +
458.9544 = 8470.21

Quarter 3
Profits = 6688.363 + 0 - 217.8054 + 0 + (0.0382462 × 12000) = 6688.363 - 217.8054 +
458.9544 = 6929.51

Quarter 4

4
Profits = 6688.363 + 0 + 0 + 183.8564 + (0.0382462 × 12000) = 6688.363 + 183.8564 +
458.9544 = 7331.17

5
 Question 2 (30)

The following table gives quarterly data on the sale of mutual fund shares by the mutual fund
industry for the period 1968-1973.

Quarter
Year I II III IV
1968 1564 1654 1607 1994
1969 2129 1658 1428 1503
1970 1381 1039 975 1230
1971 1304 1288 1108 1446
1972 1398 1176 1099 1219
1973 1382 888 933 1156

a) Using a calculator, compute mean sales for each quarter. (4)

Quarter I:
(1564 + 2129 + 1381 + 1304 + 1398 + 1382)/6 = 1526.33

Quarter II:
(1654 + 1658 + 1039 + 1288 + 1176 + 888)/6 = 1283.83

Quarter III:
(1607 + 1428 + 975 + 1108 + 1099 + 933)/6 = 1191.67

Quarter IV:
(1994 + 1503 + 1230 + 1446 + 1219 + 1156)/6 = 1424.67

b) Consider the following model:

𝑆𝑎𝑙𝑒𝑠𝑡 = 𝛼1 + 𝛼2𝐷2𝑡 + 𝛼3𝐷3𝑡 + 𝛼4𝐷4𝑡 + 𝑢𝑡

where:

D2 = 1 for the second quarter, 0 otherwise

D3 = 1 for the third quarter, 0 otherwise
D4 = 1 for the fourth quarter, 0 otherwise

i. Estimate the preceding regression (10)

6
 Sales = 1526.333 - 242.5D2 - 334.6667D3 - 101.6667D4

ii. How would you interpret the α’s (6)

 α₁ (constant) = 1526.333: This is the mean sales for Q1
 α₂ = -242.5: Q2 sales are on average 242.5 million less than Q1 sales
 α₃ = -334.6667: Q3 sales are on average 334.67 million less than Q1 sales
 α₄ = -101.6667: Q4 sales are on average 101.67 million less than Q1 sales

iii. What are the mean sales for each quarter? (4)

 Q1 mean sales = α₁ = 1526.333

 Q2 mean sales = α₁ + α₂ = 1526.333 - 242.5 = 1283.833
 Q3 mean sales = α₁ + α₃ = 1526.333 - 334.6667 = 1191.666
 Q4 mean sales = α₁ + α₄ = 1526.333 - 101.6667 = 1424.666

iv. Conduct a statistical test to determine if mean sales for the second, third and
fourth quarters are different from mean sales for the first quarter. (6)

𝐻0: 𝛼2= 𝛼3= 𝛼4 = 0

Using the f-test for significance:

𝐻1:𝛼1 ≠ 0

F = MSR/MSE
131880.931/90934.0417
=1.45

means are not statistically significant thus we fail to reject the null hypothesis 𝐻0.
Since 1.45 is less than the tabulated f-statistic 3.47, we conclude that there the group