Sqqs2083 Sampling Techniques Chapter 4: Stratified Sampling
Sqqs2083 Sampling Techniques Chapter 4: Stratified Sampling
Sqqs2083 Sampling Techniques Chapter 4: Stratified Sampling
1
Overview of stratified sampling:
Divide population into two or more subgroups (called strata) according to
some common characteristic
• A simple random sample is selected from each subgroup
• Samples from subgroups are combined into one
4
Example 4.1:
A small town contains a total of 1800
households. How to conduct a survey using
stratified random sampling?
Solution:
5
Example 4.2:
A restaurant chain has 60 restaurants in Kedah, 50
in Penang and 45 in Perlis. Management is
considering adding a new item to the menus. To
test the likely demand for this item, it was
introduced on the menus of random samples of
12 restaurants in Kedah, 10 in Penang, and 9 in
Perlis. Conduct the survey by using stratified
random sampling.
6
Solution:
7
• Let random samples of ni individuals be taken from
strata containing Ni individuals (i = 1, 2, . . ., L)
• Let L K
N
i 1
i N and n
i 1
i n
8
An unbiased estimator for the variance of the overall
population mean is
K
1
σ̂ 2x st 2
N
i σ̂ x i
N
j1
2 2
where σ̂ 2x s i2 (N i n i )
i
ni Ni
1 L
N n s 2
hence:
σ̂ 2
x st 2
N
N 2
i
i
Ni
i
n
i
i 1 i
9
An unbiased estimator for the of the overall
population proportion and variance of stratum
are:
L
1 1
pˆ st N1 pˆ 1 N 2 pˆ 2 N L pˆ L N pˆ i i
N N i 1
1 L 2 N i ni pˆ i qˆi
ˆ 2
pst
1
2 2 2
2 N1 ˆ pˆ1 N 2 ˆ pˆ 2 N Lˆ pˆ L 2 N i
N N i 1 Ni ni 1
10
• Suppose that random samples of nj individuals
from strata containing Nj individuals (j = 1,
2, . . ., K) are selected and that the quantity to
be estimated is the population total, Nμ
• An unbiased estimation procedure for the
population total Nμ yields the point estimate
L
ˆ Nx st N i x i
i 1
11
Example 4.3:
Refer to example 4.2, estimate the average number, total population and variance
of weekly orders per restaurant. Given the sample means and standard deviation for
the numbers of orders received for this item per restaurant in the three state in a
week:
xkedah 21.2 skedah 12.8
x penang 13.3 s penang 11 .4
x perlis 26.1 s perlis 9.2
Then if from each state, 9 customers from Kedah, 8 from Penang and 7 from Perlis
like the new menus. Find the population mean, variance mean, proportion and
variance of stratum proportion.
12
N kedah 60 N penang 50 N perlis 45
nkedah 12 n penang 10 n perlis 9
13
Solution:
14
Solution:
15
Exercise 4.1
A small town contains of 1800 households. The town is
divided into three districts, containing 820, 540, 440
households respectively. A stratified random sample of 300
households containing 120, 90, and 90 households,
respectively from these three districts. Sample members
were asked to estimate their total energy bills for the winter
months. The respective sample means were RM290, RM352,
and RM427. The respective sample standard deviation were
RM47, RM61, and RM93. Estimate the mean winter energy
bill, total population of winter energy bill, and the variance of
energy bill for all household in the town.
16
N1 820 N 2 540 N 3 440
n1 120 n2 90 n3 90
x1 290 x2 352 x3 427
s1 47 s2 61 s3 93
N 1800
17
Solution:
18
Solution:
19
Determining Sample Size
• The sample size is directly related to the size of
the variance of the population estimator
• If the researcher sets the allowable size of the
variance in advance, the necessary sample size
can be determined
20
PROPORTIONAL ALLOCATION: Sample Size
• One way to allocate sampling effort is to make the
proportion of sample members in any stratum the same as
the proportion of population members in the stratum
• If so, for the jth stratum,
nj Nj
n N
22
Solution:
23
Exercise 4.2:
Refer to exercise 4.1, suppose we want to select
the sample size n = 400. Find the approximation
of nj for estimating means and total population
using proportional allocation.
24
Solution:
25
OPTIMAL ALLOCATION
To estimate an overall population mean or total
and if the population variances in the individual
strata are denoted σj2 , the most precise
estimators are obtained with optimal allocation
• The sample size for the jth stratum using
optimal allocation is n j N jσ j n
K
N σ
i 1
i i
26
• To estimate an overall population mean or total
and if the population variances in the individual
strata are denoted σj2 , the most precise
estimators are obtained with optimal allocation
• The sample size for the jth stratum using
optimal allocation is n j N jσ j n
K
N σ
i 1
i i
27
• To estimate the overall population proportion,
estimators with the smallest possible variance
are obtained by optimal allocation
• The sample size for the jth stratum for
population proportion using optimal allocation
is N j Pj (1 Pj )
nj K
n
N
i 1
i Pi (1 Pi )
28
• Example 4.5:
Refer to example 4.2, we wish to choose the
sample size n = 30. Find the approximation of n
for estimating means and total population using
optimum allocation. Given
σkedah = 13, σpenang = 11, and σperlis = 9
29
Solution:
30
Solution:
31
Exercise 4.3
Refer to exercise 4.1, if we wish to take 300
household of the total. Find the approximation
of nj estimating means and total population
using optimal allocation. Given
σ1 = 50, σ2 = 60, and σ3 = 90
32
Solution:
33
NON PROPORTIONAL ALLOCATION
• Selecting n for estimating the population Means and
Total
Suppose we specify that the estimate sample mean
for stratified sample should lie within B units of the
population mean, with probability approximately
equal to 0.95. symbolically;
2
B
2 2
x st B or 2
x st
4
34
• Let ni = nai i = 1, 2, 3, …, L
where ai is a allocation fraction for stratum i
• Hence approximation n required to estimate μ
N
L 2 2
is: i i
ai
n i 1
L
N 2 D N i i2
i 1
B2
D
4
35
and approximation n required to estimate τ is:
N i2 i2
L
ai
n i 1
L
N D N i
2
i
2
i 1
B2
D 2
4N
36
Example 4.4:
37
• Solution: