Final Exam Problems - Sampling Theory | STAT 440, Exams of Survey Sampling Techniques

Material Type: Exam; Class: Sampling Theory; Subject: Statistics and Probability; University: University of Maryland; Term: Fall 1997;

Typology: Exams

Pre 2010

Uploaded on 05/13/2008

koofers-user-60b
koofers-user-60b 🇺🇸

5

(2)

10 documents

1 / 2

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
STAT
440
FINAL
EXAM
FALL,
1997
Instructions:
Work
any
three
problems.
Show
all
work
related
to
your
solutions.
You
may
use
calculators
or
a
3
x
5
card
with
formulas.
1.
In
a
population
of
N
elements,
positive
variables
x,-
and
y,'
are
de-
fined
for
each
element,
and
the
population
total
X
is
known.
It
is
proposed
to
estimate
the
population
total
Y
using
the
ratio
estimator
YR
=
X(y/x).
However,
the
sampling
plan
is
to
choose
a
sample
s
of
n
elements
with
prob-
ability
c^ies
x
i-
(a)
Show
how
to
find
c
so
that
YR
is
unbiased.
(b)
Show
that
s
can
be
selected
with
probability
c£,-
es
x,-
as
follows:
(i)
choose
a
first
element
with
probability
x,-/X;
(ii)
choose
n
I
additional
elements
by
simple
random
sampling
with-
out
replacement
from
the
remaining
N
1
elements.
2.
A
population
consists
of
N
clusters
with
M,-
elements
in
cluster
i.
The
total
number
of
elements
in
the
population,
MO,
is
unknown.
A
simple
random
sample
of
n
clusters
is
selected,
and
the
y
value
is
observed
on
each
element
of
the
selected
clusters.
The
estimator
\r
£—*l£s
eft
Y
R
=
^ rr
is
used
to
estimate
the
population
mean
per
element
N
Mi
where
y;
is
the
total
in
cluster
i.
Show
that
VarY#
is
approximately
equal
n
M
0
2
TV
-
1
and
propose
an
estimator
for
this
expression.
pf2

Partial preview of the text

Download Final Exam Problems - Sampling Theory | STAT 440 and more Exams Survey Sampling Techniques in PDF only on Docsity!

STAT 440 FINAL EXAM FALL, 1997

Instructions: Work any three problems. Show all work related to your solutions. You may use calculators or a 3 x 5 card with formulas.

  1. In a population of N elements, positive variables x,- and y,' are de- fined for each element, and the population total X is known. It is proposed to estimate the population total Y using the ratio estimator YR = X(y/x). However, the sampling plan is to choose a sample s of n elements with prob- ability c^ies x i-

(a) Show how to find c so that YR is unbiased.

(b) Show that s can be selected with probability c£,- es x,- as follows:

(i) choose a first element with probability x,-/X; (ii) choose n — I additional elements by simple random sampling with- out replacement from the remaining N — 1 elements.

  1. A population consists of N clusters with M,- elements in cluster i. The total number of elements in the population, MO, is unknown. A simple random sample of n clusters is selected, and the y value is observed on each element of the selected clusters. The estimator

\r YR = ^ rr £—*l£s eft

is used to estimate the population mean per element

N Mi

where y; is the total in cluster i. Show that VarY# is approximately equal

n M0 2 TV - 1

and propose an estimator for this expression.

  1. A simple random sample of 40 countries was selected from a population of 124 countries. Among the available data are x,-, y,-, i = l,...,n, the populations of the sampled countries in 1980 and 1983, respectively. The sample data were reduced as follows:

y = 42.49; sj = 26,115.51; sxy = 25,582.08;

N x- 40.89; 4 = 25,061.00; X = £z; = 4, 308.1. ! = 1 Populations are measured in millions of people.

(a) Compute the difference estimator for Y and estimate its variance.

(b) It is thought that the yearly rate of increase of the total population of the 124 countries is between 1% and 2%. If this is true, suggest a better estimator of Y than the one in (a).

(c) Compute the ratio estimator of Y and estimate its variance.

  1. A simple random sample of n units from a population of size N yields data (x t-,j/i), i 6 s. The population quantity X is known. Let

n

Show that the statistic

YHR = rA +^ v-^ N - 1^ n^ -(y,^ - rx)^ 1\ Tl — 1

is an unbiased estimator of Y.