Chapter Goals

After completing this chapter, you should be

Engineering Statistics able to:
• Define the concept of sampling error
• Determine the mean and standard deviation for
Chapter 6 the sampling distribution of the sample mean, x
Introduction to • Describe the Central Limit Theorem and its
Sampling Distributions importance
• Apply sampling distributions for x .
1 of 32 2 of 32

Sampling Error Calculating Sampling Error

• Sampling Error:
• Sample Statistics are used to estimate The difference between a value (a statistic)
Population Parameters computed from a sample and the
corresponding value (a parameter) computed
ex: x is an estimate of the population mean, μ
from a population.
• Problems:
– Different samples provide different estimates of the Example: (for the mean)
population parameter
– Sample results have potential variability, thus sampling
Sampling Error = x - μ
error exists. where:
x = sample mean
μ = population mean.
3 of 32 4 of 32
 Review Example
• Population mean: Sample mean:
If the population mean is μ = 99.2 degrees

μ=
åx i
x=
åx i
and a sample of n = 5 temperatures yields a
sample mean of x = 98.6 degrees, then the
N n sampling error is:
where:
μ = Population mean
x - μ = 98.6 - 99.2 = -0.6 degrees
x = sample mean
xi = Values in the population or sample
N = Population size
n = sample size.
5 of 32 6 of 32

Sampling Errors Sampling Distribution

• Different samples will yield different sampling • A sampling distribution is a

errors
distribution of the possible values of
• The sampling error may be positive or negative
( x may be greater than or less than μ) a statistic for a given sample size
• The expected sampling error decreases as the selected from a population.
sample size increases.

7 of 32 8 of 32
 Developing a Developing a
Sampling Distribution Sampling Distribution (continued)

Summary Measures for the Population Distribution:

• Assume there is a population …
A B C D
μ=
åx i
P(x)
• Population size N=4 N .3
• Random variable, x, 18 + 20 + 22 + 24 .2
= = 21
is age of individuals 4 .1
• Values of x: 18, 20, 0
σ=
å (x i - μ)2
= 2.236 18 20 22 24 x
22, 24 (years). N A B C D
Uniform Distribution

9 of 32 10 of 32

Developing a Developing a
Sampling Distribution (continued) Sampling Distribution (continued)
Now consider all possible samples of size n=2 Sampling Distribution of All Sample Means
st nd
1 2 Observation
Obs 16 Sample 16 Sample Means Sample Means
18 20 22 24
Means Distribution
18 18,18 18,20 18,22 18,24 1st 2nd Observation
1st 2nd Observation
20 20,18 20,20 20,22 20,24 Obs 18 20 22 24 P(x)
Obs 18 20 22 24
.3
22 22,18 22,20 22,22 22,24 18 18 19 20 21 18 18 19 20 21
.2
24 24,18 24,20 24,22 24,24 20 19 20 21 22 20 19 20 21 22
.1
16 possible samples 22 20 21 22 23 22 20 21 22 23
(sampling with 0 _
replacement)
24 21 22 23 24 24 21 22 23 24 18 19 20 21 22 23 24 x
11 of 32 (no longer uniform) 12 of 32
 Developing a Comparing the Population with its
Sampling Distribution (continued) Sampling Distribution
Summary Measures of this Sampling Distribution:
Population Sample Means Distribution
n=2
μx =
åx i
=
18 + 19 + 20 + ! + 24
= 21 μ = 21
N=4
σ = 2.236 μx = 21 σ x = 1.58
N 16
P(x) P(x)
.3 .3
σx =
å (x i - μ x )2
.2 .2
N
.1 .1
(18 - 21)2 + (19 - 21)2 + ! + (24 - 21)2
= = 1.58 0 0 _
16 18 20 22 24 x 18 19 20 21 22 23 24 x
A B C D
13 of 32 14 of 32

If the Population is Normal z-value for Sampling Distribution

of x
(THEOREM 6-1)
• Z-value for the sampling distribution of x :
If a population is normal with mean μ and
(x - μ)
standard deviation σ, the sampling distribution z=
σ
of x is also normally distributed with: n
where: x = sample mean
σ
μx = μ and σx = μ = population mean
n σ = population standard deviation
n = sample size
15 of 32 16 of 32
 Finite Population Correction Sampling Distribution Properties

• Apply the Finite Population Correction if: Normal Population

– the sample is large relative to the population • μx = μ Distribution

(n is greater than 5% of N)
and… (i.e. x is unbiased ) μ x
– Sampling is without replacement Normal Sampling
Distribution
(x - μ) (has the same mean)
Then z=
σ N-n
μx
n N -1 x
17 of 32 18 of 32

Sampling Distribution Properties If the Population is not Normal

(continued)
• We can apply the Central Limit Theorem:
• For sampling with replacement:
– Even if the population is not normal,
As n increases, Larger
σ x decreases sample size – …sample means from the population will be
approximately normal as long as the sample size is
large enough
Smaller
– …and the sampling distribution will have:
sample size
σ
μx = μ and σx =
n
μ x
19 of 32 20 of 32
 Central Limit Theorem If the Population is not Normal
(continued)

Population Distribution
the sampling Sampling distribution
As the n↑
distribution properties:
sample
becomes Central Tendency
size gets
almost normal μx = μ
large μ x
regardless of
enough… Sampling Distribution
shape of Variation
σ (becomes normal as n increases)
population σx = Larger
n Smaller
sample size
sample
size
(Sampling with
replacement)
x
21 of 32 μx x
22 of 32

How Large is Large Enough?

Example
• For most distributions, n > 30 will give a • Suppose a population has mean μ = 8 and
sampling distribution that is nearly standard deviation σ = 3. Suppose a random
normal sample of size n = 36 is selected.

• For fairly symmetric distributions, n > 15 • What is the probability that the sample mean is
• For normal population distributions, the between 7.8 and 8.2?
sampling distribution of the mean is
always normally distributed.

23 of 32 24 of 32
 Example (continued)
Example (continued)
Solution: Solution (continued):
æ ö
• Even if the population is not normally ç 7.8 - 8 μx -μ 8.2 - 8 ÷
distributed, the central limit theorem can be P(7.8 < μ x < 8.2) = Pç < < ÷
ç3 σ 3 ÷
used (n > 30) è 36 n 36 ø

• … so the sampling distribution of x is = P(-0.4 < z < 0.4) = 0.3108

approximately normal Population Sampling Standard Normal
Distribution Distribution Distribution
• … with mean μx = 8 ???
.1554
+.1554
? ??
σ 3 ?
• …and standard deviation σ x = = = 0.5 ??
? ?
?
Sample Standardize
n 36
μ=8 x 7.8
μx = 8 8.2 x -0.4
μz = 0 0.4 z
25 of 32 26 of 32

• Central Limit Theorem sangat penting dalam

• Misalkan, jumlah kunjungan di Puskesmas
statistika inferensial karena dengan teorema
ini memungkinkan kita untuk menafsirkan per tahun berdistribusi miring ke kiri
dengan rata-rata 60 orang per hari dan
parameter populasi dari sampel tanpa harus
standar deviasi sebesar 20. Bila kita ambil
mengetahui distribusi populasi.
sampel sebesar 30 hari buka maka berapa
• Dalam teorema ini diketahui bahwa untuk probabilitas jumlah kunjungan lebih dari 65
pendekatan ke distribusi normal, distribusi
orang?
rata-rata sampel tidak memerlukan sampel
• Jawab:
yang besar. Dengan sampel sebesar 30 telah
terjadi pendekatan ke distribusi normal.
Business Statistics: A Decision-Making Chap 6-27 Business Statistics: A Decision-Making Chap 6-28
Approach, 6e © 2005 Prentice-Hall, Inc. 27 of 32 Approach, 6e © 2005 Prentice-Hall, Inc. 28 of 32
 Jawab :
Pertama-tama, hitunglah kesalahan baku rata-rata kemudian hitung
nilai Z untuk 65 orang. Chapter Summary
• Discussed sampling error
• Introduced sampling distributions
• Described the sampling distribution of the mean
– For normal populations
– Using the Central Limit Theorem
• Calculated probabilities using sampling
distributions
• Discussed sampling from finite populations.

Business Statistics: A Decision- Chap 6-29

Making Approach, 6e © 2005 29 of 32 30 of 32
Peluang Z=1,37 adalah 0,085 atau 8,5%.
Prentice-Hall, Inc.

Thank You

31 of 32