Scribd
Upload
202 views

Answers To Homework 1: Exercise 1.1

The document provides answers to homework questions related to an economics course. It includes explanations of concepts like independent and dependent variables in regression models. It also works through examples of hypothesis testing, confidence intervals, and other statistical analyses. Key terms and formulas are defined throughout as the questions are addressed step-by-step.

Uploaded by

Osman Furkan Abbasoglu
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
202 views

Answers To Homework 1: Exercise 1.1

The document provides answers to homework questions related to an economics course. It includes explanations of concepts like independent and dependent variables in regression models. It also works through examples of hypothesis testing, confidence intervals, and other statistical analyses. Key terms and formulas are defined throughout as the questions are addressed step-by-step.

Uploaded by

Osman Furkan Abbasoglu
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

ECON 414 Answers to Homework 1

Answers to Homework 1
1 Chapter 1
Exercise 1.1
We would expect ice cream consumption to depend on its price and how hot the day is.
Income might also be important, but less so than the above variables. Thus, a possibe
specication is
Demand =
1
+
2
Income +
2
Price +
4
Temp +u
where
Demand is the weekly consumption of ice cream (say, in gallons).
Price is the price per gallon.
Income is the average weekly income.
Temp is the average of the daily high temperatures for the week.
u is the error term.
As temperature goes up, we would expect more people to buy ice cream and hence we would
expect
4
to be positive. If ice cream is more expensive, then demand would fall. Hence, we
would expect
3
to be negative. Finally, income eect is expected to be positive and hence

2
is likely to be positive.
If we obtain cross section data at a given point in time, that is, across individuals, then
temp and price will not vary across individuals and hence we cannot estimate the model.
Therefore, time series data are more appropriate.
Exercise 1.4
There are two types of variables in a regression model: independent and dependent. The
dependent variable in a regression model is the attribute of an observation which we are
interested in studying. An independent variable is an attribute that aects the value of the
dependent variable.
Parameters explain how independent variables aect on dependent variable. In the sense
of economic theory, a parameter is the marginal eect on the dependent variable of a unit
increase in a corresponding independent (that is, explanatory) variable.
Exercise 1.5
See p.607 in the textbook
1
ECON 414 Answers to Homework 1
2 Chapter 2
Exercise 2.5
Let X be the number of bulbs out of 20(= n) that do not germinate. And the probability of
not germinating is p = 0.2 and hence
X Bin(20, 0.2)
f
X
(x) =
20!
x!(20 x)!
0.2
x
0.8
20x
x = 0, 1, , 20
My wife will not get a refund if X 2. We therefore need P(X 2) for a Bin(20, 0.2)
From Table A.6,
P(X 2) = 1 P(X 3) = 1 0.7939 = 0.2061
Or you can calculate this probability directly.
P(X 2) = f
X
(0) +f
X
(1) +f
X
(2) = 0.2061
Exercise 2.10
First location:
E(profit) = 0.5 20000 + 0.5 (2000) = 10000 1000 = 9000
Second location:
E(profit) = 0.5 25000 + 0.5 (5000) = 12500 2500 = 10000
The company will be on the second location.
Exercise 2.15
V ar(Y bX) = V ar(Y ) +b
2
V ar(X) 2bCov(X, Y )
=
2
y
+b
2

2
x
2b
xy
In order to obtain b which minimize V ar(Y bx), take the derivative of V ar(Y bx) with
respect to b and nd b which makes the derivative to be equal to zero.
2b
2
x
2
xy
= 0
b =

xy

2
x
2
ECON 414 Answers to Homework 1
Exercise 2.16
1) Since x
i
, i = 1, 2, , n are random sample, x
i
s are independent and identically distrib-
uted (i.i.d).
E( x) = E
_
1
n
n

i=1
x
i
_
=
1
n
n

i=1
E(x
i
) =
1
n
n

i=1
=
V ar( x) = V ar
_
1
n
n

i=1
x
i
_
=
1
n
2
n

i=1
V ar(x
i
) =
1
n
2
n

i=1

2
=

2
2) For E(y), V ar(y),
E(y) = E
_
1
n
n

i=1
a
i
x
i
_
=
1
n
n

i=1
a
i
E(x
i
) =
1
n
n

i=1
a
i
=

n
n

i=1
a
i
V ar(y) = V ar
_
1
n
n

i=1
a
i
x
i
_
=
1
n
2
n

i=1
a
2
i
V ar(x
i
) =
1
n
2
n

i=1
a
2
i

2
=

2
n
2
n

i=1
a
2
i
For E(y) to be equal to ,
E(y) =

n
n

i=1
a
i
=

1
n
n

i=1
a
i
= 1

i=1
a
i
= n
3) Now recall the fact:
If a random variable X with E(X) = and V ar(X) =
2
, then a random variable
Z =
X

has
E(Z) = E
_
X

_
=
1

E(X ) =
1

(E(X) ) = 0
V ar(Z) = V ar
_
X

_
=
1

2
V ar(X ) =

2

2
= 1
We can apply this equation to x to get the desired Z. Since E( x) = and V ar( x) =

2
n
,
Z =
x

n
=

n( x )

is the desired random variable


3
ECON 414 Answers to Homework 1
Exercise 2.18
Since x
1
, x
2
, and x
3
have the same mean,
E(
1
) = 0.2E(x
1
) + 0.3E(x
2
) + 0.5E(x
3
) = 0.2 + 0.3 + 0.5 =
E(
2
) = 0.4E(x
1
) + 0.2E(x
2
) + 0.4E(x
3
) = 0.4 + 0.2 + 0.4 =
E(
3
) = 0.3E(x
1
) + 0.3E(x
2
) + 0.3E(x
3
) = 0.3 + 0.3 + 0.3 = 0.9
The rst two estimators are unbiased, while the last one is downward biased. Since x
1
, x
2
,
and x
3
are independent and identical distributed,
V ar(
1
) = 0.2
2
var(x
1
) + 0.3
2
var(x
2
) + 0.5
2
var(x
3
) = 0.38
2
V ar(
2
) = 0.4
2
var(x
1
) + 0.2
2
var(x
2
) + 0.4
2
var(x
3
) = 0.36
2
Thus
2
is more ecient than
1
.
Exercise 2.22
1. Sample means are
y =

n
i=1
y
i
n
=
107226
500
= 214.452
x =

n
i=1
x
i
n
=
24838
500
= 49.676
Sample variances are
S
2
y
=

n
i=1
(y
i
y)
2
n 1
=
1398308
499
= 2802.22
S
2
x
=

n
i=1
(x
i
x)
2
n 1
=
66398
499
= 133.062
Then, sample standard deviations are
S
y
=

2802.22 = 52.936
S
z
=

133.062 = 11.535
2. The (sample) correlation coecient is
r
xy
=

n
i=1
(x
i
x)(y
i
y)
_

n
i=1
(x
i
x)
2
_

n
i=1
(y
i
y)
2
=
194293

66398

1398308
= 0.638
3. Assume that the sample houses are drawn independently from the same population.
Before construction of the condence interval for the mean value of house, we should
4
ECON 414 Answers to Homework 1
set t

to be the point such that P(t

t
499
t

) = 0.95 where t
499
is a t-distribution
with the degree of freedom, 499. From Table A.2., t

= 1.96 (Look up the value in the


row where d.f. is and the column where 2T = 1 0.95 = 0.05). Then the condence
interval for the mean value of house is
y
_
t

s
y

n
_
= 213.452
_
1.96
52.936

499
_
= 213.452 4.645 = (209.807, 219.097)
4. The test statistics is
F
c
=
(n 2)r
2
1 r
2
=
498(0.638)
2
1 0.638
2
= 341.9
Under the null hypothesis of zero correlation, the test statistics has an F-distribution
with 1 d.f. for the numerator and 498 d.f. for the denominator. From Table A.4a we
note that F

1,498
=6.63. Since F
c
F

1,498
, we reject the null hypothesis and conclude
that there is a signicant correlation between household income and the vlaue of houses
at the level of 1%.
Exercise 2.23
We have
n = 81 x = 739.98 S
x
= 312.7
From the t-table t

= 1.993 (Look up the value in the column where 2T = 1 0.95 = 0.05


and the row where d.f. is 80. But we do not have d.f. 80 so we should interpolate between
60 and 120 d.f.)
The condence interval for the mean claim is therfore
x
_
t

S
x

n
_
= 739.98
_
1.993
312.7

81
_
= 739.98 69.246 = (670.734, 809.226)
Because this condence interval includes the value 800 in the null hypothesis, we do not reject
the null hypothesis H
0
.
Another way to test the hypotheses is to construct the test statistics,
t
c
=

n( x 800)
S
x
=
9 (739.98 800)
312.7
= 1.727
Under the null hypothesis H
0
: = 800, t
c
has the t-distribution with n 1 d.f. Since
|t
c
| < t

, we do not reject H
0
at the 5% level.
5

You might also like