Scribd
Upload
269 views2 pages

HW 1

This document contains instructions for 5 homework assignments in a statistics course. It lists statistics problems to solve from textbooks and provides additional optional problems. It clarifies a question for homework 1 and provides due dates for homework submissions throughout the semester.

Uploaded by

Bob Sanders
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
269 views2 pages

HW 1

This document contains instructions for 5 homework assignments in a statistics course. It lists statistics problems to solve from textbooks and provides additional optional problems. It clarifies a question for homework 1 and provides due dates for homework submissions throughout the semester.

Uploaded by

Bob Sanders
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
You are on page 1/ 2

Statistics 593 Homework Spring 2017

Note: The problems are taken from the Exercises in Casella and Berger (2002) unless oth-
erwise noted. For each problem, please explain your reasoning clearly. It is not acceptable
to only provide your final result.

Homework 1 (Due Wed, Feb 1):


6.9, 6.10, 6.11, 6.12, 6.15, 6.20, 6.23, 6.30, 6.31(b,c)

The question 6.31(b)(ii) should be corrected as follows. Suppose X1 , . . . , XN are N Monte


Carlo samples, where each Xj consists of an iid sample of size n from N(, 2 ). Let M j be
the sample median and X j the sample mean, both computed from Xj . Let M be the average
be the average of (X 1 , . . . , X N ). A naive estimator of the variance
of (M 1 , . . . , M N ), and X
of the sample median is then

1 j
N
v1 = (M M )2 .
N 1 j=1

The swindle estimator of the variance of the sample median is

1
N
2 )}2 .
v2 = + {M j X j (M X
n N 1 j=1

Show that the variance of v1 is approximately 2[var(M )]2 /(N 1), and the variance of v2 is
approximately 2[var(M X)]2 /(N 1).

Homework 2 (Due Wed, Feb 22):


6.37, 7.2, 7.6, 7.10, 7.12, 7.13, 7.18, 7.23, 7.24,
7.37, 7.38, 7.42, 7.44, 7.45, 7.46, 7.49, 7.52

Additional problem I:
Let X1 , . . . , Xn be an iid sample from N(, 2 ) with unknown (, 2 ).
1) Find the canonical parameters for N(, 2 ) as an exponential family.
2) Derive the maximum likelihood estimators of (, 2 ) by solving the score equation in
terms of the canonical parameters.

Additional problem II:


Let X be an observation from N (, 1) with unknown . Suppose that the prior on is
N(0, 2 ) with a fixed > 0.
1) Show that the maximum likelihood estimator of is = X.
2) Find the posterior mean of , denoted by , as a point estimator.
3) Find the bias, variance, and mean squared error of . [Continue to next page]

1
4) Determine the condition, depending on and , such that the mean square error of is
no greater than that of .

Additional problem III:


Let (X1 , . . . , Xn ) be an iid sample from Uniform (0, ) with unknown > 0. Consider the
n+1
estimator = n
X(n) .
1) Show that is unbiased for .
2) Show that
2
var () = .
n(n + 2)

Homework 3 (Due Wed, March 29):


7.62, 7.65, 8.5, 8.6, 8.7(a), 8.11, 8.12, 8.15, 8.22, 8.23,
8.28, 8.29, 8.31, 8.40, 8.41, 8.47, 8.52, 8.53

Additional problem I:
Let (X1 , . . . , Xn ) be an iid sample from N(x , x2 ) and, independently, (Y1 , . . . , Ym ) be an iid
sample from N(y , y2 ), where x2 and y2 are known. Consider testing

H0 : x = y = 0,
H1 : x = 0 or y = 0.

1) Find a likelihood ratio test (LRT) rejection region and a union-intersection test (UIT)
rejection region to achieve size .
For the remaining questions, set = 10%.
2) Compute and plot the power function for LRT and UIT rejection regions along the direc-
tion x = 0 and y R. You may use R or other language.
3) Compute and plot the power function for LRT and UIT rejection regions along the direc-
tion x = y R.
4) Compute and plot the power function for LRT and UIT rejection regions for (x , y ) R2 .
Choose a plotting method that you think most appropriate.

Homework 4 (Due Wed , Apr 12):


8.26, 9.3, 9.4, 9.6, 9.13, 9.21, 9.23, 9.25, 9.29, 9.32, 9.41, 9.45, 9.48(ab)

Homework 5 (Due Wed, May 3):


10.3, 10.9, 10.10(ab), 10.12, 10.29, 10.31, 10.35(b), 10.37(b),
10.17(d) [Verify the normality result, i.e, complete tedious matrix calculations.]

You might also like