VERSION A

BSc-III
Econometrics I (ECO 204)
Quiz 1
Spring 2022
Azam Chaudhry Aimal Tanvir
SOLUTION
Name: _________________ Section: _______
Total Marks: 80 marks
Question 1 (10 marks):

Below are the results of a regression of life expectancy in years (lifeex) on number of televisions per
hundred people (tv).

a. Construct the estimated regression equation using the above table. (4 marks)

̂ 𝒊 = 𝟓𝟕. 𝟑𝟏 + 𝟎. 𝟒𝟏𝟔 𝒕𝒗𝒊

b. Interpret the coefficient on tv. (2 marks)

c. What is meant by the R-squared and why is it used? The value of the R-squared is 0.5370.
Provide an interpretation for the R-squared. (4 marks)

R-squared explains how much of the sample variation in the dependent variable is being explained by
the independent variable. It is used to see how well the model fits the actual data points. The value of
the R-squared ranges from 0 to 1; where a value close to 0 represents a poor fit and a value close 1
represents a near-perfect fit. The estimated value of the R-squared is 0.5370, suggesting that 53.70%
of the variation in the life expectancy is being explained by the number of televisions.

Page 1 of 4
 VERSION A

Question 2 (40 marks):

Suppose you are interested in estimating the effect of hours spent in an SAT preparation course
(hours) on total SAT score (sat). The population is all college-bound high school seniors for a
particular year. Suppose that you have randomly collected information on 6 students from the
population on the number of hours they choose to spend in the preparation course and their SAT
score.

a. Write down the relevant population regression equation for the example above by making use
of the variable names hours and sat. List at least two factors contained in the error term. Are
these likely to have positive or negative correlation with hours? (8 marks)

𝒔𝒂𝒕𝒊 = 𝜷𝟎 + 𝜷𝟏 𝒉𝒐𝒖𝒓𝒔𝒊 + 𝝁𝒊

There are a number of factors that could be correlated with hours such as intelligence, high
school GPA, family income, time spent on other activities, etc. You had to mention any two and
explain with the help of economic reasoning its correlation with hours.

b. In the equation from part (a), what is the interpretation of 𝛽0 ? (4 marks)

𝜷𝟎 is the y-intercept in the regression equation and it gives us the average SAT score of
students who did not prepare for the test i.e., hours=0.

You have collected the following data from the students in the population:

student hours (X) sat (Y) (x-x̅) (x-x̅)2 (y-y̅) (x-x̅)(y-y̅) ŷ (y-ŷ)

1 10 1020 -8.167 66.694 -60.833 496.806 836.93 183.07

2 35 1550 16.833 283.361 469.167 7897.639 1583.58 -33.58

3 22 1275 3.833 14.694 194.167 744.306 1195.322 79.678

4 7 600 -11.167 124.694 -480.833 5369.306 747.332 -147.332

5 18 1090 -0.167 0.028 9.167 -1.528 1075.858 14.142

6 17 950 -1.167 1.361 -130.833 152.639 1045.992 -95.992

18.167 1080.833 490.833 14659.167 -0.014

slope 29.866

intercept 538.270

c. Estimate the relationship between sat and hours using OLS; that is, obtain the intercept and
slope estimates. (10 marks)
̂𝒊 = 𝟓𝟑𝟖. 𝟐𝟕𝟎 + 𝟐𝟗. 𝟖𝟔𝟔 𝒉𝒐𝒖𝒓𝒔𝒊
𝒔𝒂𝒕

Page 2 of 4
 VERSION A
d. Comment on the direction of the relationship. (4 marks)
Interpret the slope coefficient

e. How much higher is the SAT score predicted to be if the number of hours increase by 6? (4
marks)
Plug 6 into the estimated equation in part c and solve.

f. Compute the fitted values and residuals for each observation, and verify that the residuals
(approximately) sum to zero. (10 marks)

Plug in values of hours into the following equation and compute the fitted values:
̂𝒊 = 𝟓𝟑𝟖. 𝟐𝟕𝟎 + 𝟐𝟗. 𝟖𝟔𝟔 𝒉𝒐𝒖𝒓𝒔𝒊
𝒔𝒂𝒕
Use the following formula to compute the residuals:
𝜇̂i = yi − ŷi
Make sure you clearly state your working for this part, otherwise, marks will be deducted.

Question 3 (30 marks):

Using observations for 100 families on annual income and savings (both measured in dollars), the
following equation is obtained:

̂ 𝐢 = −𝟏𝟐𝟒. 𝟖𝟒 + 𝟎. 𝟕𝟖𝟒 𝐢𝐧𝐜𝐨𝐦𝐞𝐢

Following are the estimated equations using alternative functional forms for this relationship:
̂ 𝐢 ) = −𝟎. 𝟖𝟗 + 𝟎. 𝟗𝟎𝟐 𝐢𝐧𝐜𝐨𝐦𝐞𝐢
𝐥𝐧(𝐬𝐚𝐯𝐢𝐧𝐠 𝐄𝐪𝐮𝐚𝐭𝐢𝐨𝐧 (𝟐)

̂ 𝐢 = −𝟎. 𝟖𝟗 + 𝟎. 𝟗𝟎𝟐 𝐥𝐧 (𝐢𝐧𝐜𝐨𝐦𝐞𝐢 )

̂ 𝐢 ) = −𝟎. 𝟖𝟗 + 𝟎. 𝟗𝟎𝟐 𝐥𝐧 (𝐢𝐧𝐜𝐨𝐦𝐞𝐢 )

a. Interpret the parameter on income (or log(income)) in each of the equations given above. (15
marks)
On average, if annual income increases by one dollar, then savings increase by 0.784 dollars,
ceteris paribus.
On average, if annual income increases by one dollar, then savings increase by 90.2%, ceteris
paribus.

Page 3 of 4
 VERSION A
On average, if annual income increases by one percent, then savings increase by 0.00902 dollars,
ceteris paribus.

On average, if annual income increases by one percent, then savings increase by 0.902 percent,
ceteris paribus.

b. What is the predicted saving with an income of $7 with each of the models? (15 marks)

Plug 7 into the estimated equations above and solve.

Solution
EQ 1 -119.352
EQ 2 5.424 226.78
EQ 3 0.865211
EQ 4 0.865211 2.38

Page 4 of 4