Principles of Econometrics Tutorial 4 answers 1

Tutorial 4 answers

1. A random sample of 16 elements is drawn from a population. The following results were
recorded (where n = 16):
n n
∑ Xi = 800; ∑ Xi2 = 40, 540
i=1 i=1

(a) The sample mean is given by

1 n 800
x= ∑ xi = = 50.
n i=1 16

(b) The sample variance is given by

!
n
1 1
540
s2 = ∑ xi2 − nx2 = 40540 − 16 × 502 = = 36.
n−1 i=1 15 15

36
(c) If X ∼ N(45, 36), σ2X = 16 so σX = 64 = 23 . The probability of obtaining a sample
mean greater than the 50 is
     
50 − µX 5 10
P(X > 50) = P Z > =P Z> =P Z> = 0.0004.
σX 1.5 3

(d) Now, σX2 = 36 6 3

64 so σX = 8 = 4 . The probability of obtaining a sample mean greater
than the 50 is
     
50 − µX 5 20
P(X > 50) = P Z > =P Z> =P Z> = 1.3 × 10−11 .
σX 0.75 3

With a greater sample size, our estimate of the mean of X is more precise, so sub-
stantial differences between the sample and population mean are less likely.

2. The random variable X.

(a) The mean and variance of X are given by

n
µ = E(X) = ∑ xi f (xi ) = 0.5 + 2(0.3) + 3(0.2) = 1.7
i=1
σ = E(X − µ) = E(X 2 ) − µ2 = 0.5 + 4(0.3) + 9(0.2) − 1.72 = 3.5 − 2.89 = 0.61.
2 2
Principles of Econometrics Tutorial 4 answers 2

(b) The mean and variance of X are given by:

µX = µ = 1.7
σ2 0.61
σ2X = = ,
n n
where n is the sample size.
(c) A random sample of size n = 2. The samples and their associated probabilities are:

Si {1,1} {1,2} {1,3} {2,1} {2,2} {2,3} {3,1} {3,2} {3,3}

P(Si ) 0.25 0.15 0.1 0.15 0.09 0.06 0.1 0.06 0.04
x 1 1.5 2 1.5 2 2.5 2 2.5 3
[The bottom row (x) is used for part (2e).]
(d) See above.
(e) The distribution of the sample mean, x:

x 1 1.5 2 2.5 3
f (x) 0.25 0.3 0.29 0.12 0.04
(f) The mean and variance of x:

µx = E(x) = 0.25 + 1.5(0.3) + 2(0.29) + 2.5(0.12) + 3(0.04)

= 0.25 + 0.45 + 0.58 + 0.3 + 0.12 = 1.7 = µ
σsx = E(x − µx )2 = E(x2 ) − µx2
= 0.25 + 2.25(0.3) + 4(0.29) + 6.25(0.12) + 9(0.04) − 1.72
= 0.25 + 0.675 + 1.16 + 0.75 + 0.36 − 2.89
σ2
= 3.195 − 2.89 = 0.305 =
2
These answers match our answers for part (2b).

3. The sample mean, X.

(a) By the Central Limit Theorem, X = 1n ∑ni=1 Xi is normally distributed if {Xi }ni=1 (i.e.
X1 , X2 , . . . , Xn ) are obtained from a random sample (as the sample size tends to in-
finity).
Principles of Econometrics Tutorial 4 answers 3

(b) The expected value of the sample mean is given by

!
1 n
E(X) = E ∑ Xi
n i=1
!
n
1
= E ∑ Xi
n i=1
1 n
= ∑ E(Xi)
n i=1
1 n
= ∑µ=µ
n i=1

(c) The expected value of the estimator X̂ is given by

!
n
E(X̂) = E ∑ kiXi
i=1
n
= ∑ ki E (Xi )
i=1
n
= ∑ ki µ
i=1
n
= µ ∑ ki .
i=1

We therefore require that ∑ni=1 ki = 1.

(d) The variance of X is given by
!2
1 n
E(X − µ)2 = E ∑ Xi − µ
n i=1
 2
1
=E (X1 + X2 + X3 ) − µ
3
 2
1
=E ((X1 − µ) + (X2 − µ) + (X3 − µ))
3
1
= E (X1 − µ)2 + (X2 − µ)2 + (X3 − µ)2 (by independence of the x’s)


9
1 1
= E(Xi − µ)2 = σ2 .
3 3
Principles of Econometrics Tutorial 4 answers 4

(e) The variance of X̂ is given by

E(X̂ − µ)2 = E (k1 X1 + k2 X2 + k3 X3 − µ)2

= E (k1 (X1 − µ) + k2 (X2 − µ) + k3 (X3 − µ))2
= E (k1 (X1 − µ))2 + (k2 (X2 − µ))2 + (k3 (X3 − µ))2 (by independence of the x’s)

= k12 E(X1 − µ)2 + k22 E(X2 − µ)2 + k32 E(X3 − µ)2

= (k12 + k22 + k32 )σ2
 
1 1 1 3
= + + σ2 = σ2 .
16 4 16 8

(f) Both X and X̂ are unbiased estimators of µ. However, X has a lower variance than
X̂.