Scribd
Upload
11 views2 pages

DOTE2011 In-Class Exercise Solutions

The document provides solutions to various statistical exercises involving confidence intervals, coin flip probabilities, hypothesis testing, and variance in linear regression. It includes specific examples and the correct answers for each question, along with explanations for why certain options are correct or incorrect. The exercises cover concepts such as normal distribution, Type I error, and least squares estimation.

Uploaded by

henrytsang1835
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
11 views2 pages

DOTE2011 In-Class Exercise Solutions

The document provides solutions to various statistical exercises involving confidence intervals, coin flip probabilities, hypothesis testing, and variance in linear regression. It includes specific examples and the correct answers for each question, along with explanations for why certain options are correct or incorrect. The exercises cover concepts such as normal distribution, Type I error, and least squares estimation.

Uploaded by

henrytsang1835
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

DOTE2011 In-Class Exercise Solutions

1. Given a set of random samples {x1 , . . . , xn } with xi ∼ N (µ, σ 2 ) for each i = 1, . . . , n.


Suppose σ is known. Which of the following interval(s) is/are interval estimate(s) of µ
with 95% confidence level? Circle the correct answer(s).
(A) (x̄ − z0.05 √σn , ∞)
(B) (x1 − z0.025 σ, x1 + z0.025 σ)
(C) (x̄ − z0.01 √σn , x̄ + z0.04 √σn )
(D) (−∞, x̄ + z0.05 √σn )
Solution: A, B, C, D
Note that
σ
P (µ > x̄ − z0.05 √ ) = P (x1 − z0.025 σ < µ < x1 + z0.025 σ) = 0.95
n
σ σ σ
P (x̄ − z0.01 √ < µ < x̄ + z0.04 √ ) = P (µ < x̄ + z0.05 √ ) = 0.95
n n n

2. Flip a coin 3 times and observe the sequence of heads and tails. For example, {T T H}
means 2 tails (T ) in the first two flips and 1 head (H) in the last flip. Which of the
following statement describes the event {T HH, HT H, HHT }?
(A) Exactly one head
(B) At most two tails
(C) At least one head
(D) None of the above describes the event.
Solution: D
(A) is not correct as we have some outcomes having more than one head. (B) is not
correct as {T T H}, which is a case of “At most two tails”, is not included in the event.
(C) is not correct as {T T H}, which is a case of “At least one head”, is not included
in the event.
3. Given a set of random samples {x1 , . . . , xn } with xi ∼ N (µ, σ 2 ) for each i = 1, . . . , n.
Suppose σ is known. We want to test H0 : µ = 0 versus H1 : µ > 0. Which of the
following rejection criteria having Type I error 0.05? Circle the correct answer(s).

1
x̄obs
(A) P (Z > √ )
σ/ n
< 0.05, where Z ∼ N (0, 1) and x̄obs is the observed sample mean
x̄√
(B) σ/ n
> z0.05
x̄√
(C) σ/ n
< −z0.05
x̄√ x̄√
(D) σ/ n
> z0.025 or σ/ n
< −z0.025

Solution: A, B, C, D
(A) is comparing p-value with the significant level 0.05. And significant level is the
probability of Type I error.
Note that the definition of the probability of Type I error is the probability of satisfying
the rejection criterion under H0 . The probability of (B) under H0 : µ = 0 is 0.05. The
same for (C) and (D).

4. In the simple linear regression model Yi = α1 + β(xi − x̄) + ϵi for i = 1, . . . , n,


Pwhere
ϵi ∼ N (0, σ ) independently. The least squares estimator of α1 is αˆ1 . x̄ = n ni=1 xi ,
2 1

what is variance of αˆ1 .

σ2
(A) V ar(αˆ1 ) = .
n
 Pn 2

1 ( i=1 (x i − x̄))
(B) V ar(αˆ1 ) = σ 2 + Pn 2
.
n i=1 (xi − x̄)
x̄2
 
2 1
(C) V ar(αˆ1 ) = σ 1 + + Pn 2
.
n i=1 (xi − x̄)
x̄2
 
2 1
(D) V ar(αˆ1 ) = σ + Pn 2
.
n i=1 (xi − x̄)

Solution: A, B
Let wi = xi − x̄, then Yi = α1 + βwi + ϵi and hence

w̄2
 
2 1
V ar(αˆ1 ) = σ + Pn 2
,
n i=1 (wi − w̄)
Pn
where w̄ = n1 P i=1 wi . Note that w̄ = 0 and hence (A) is correct.
Also note that ni=1 (xi − x̄) = nw̄ = 0, hence (B) is also correct.

You might also like