Sampling Methods: - Reasons To Sample

Sampling Methods
• REASONS TO SAMPLE
1. TO OBSERVE THE WHOLE POPULATION WOULD BE TIME
CONSUMING AND VERY EXPENSIVE.
2. THE PHYSICAL IMPOSSIBILITY OF CHECKING ALL ITEMS IN
THE POPULATION.
3. THE DESTRUCTIVE NATURE OF SOME TESTS.
4. THE SAMPLE RESULTS ARE ADEQUATE TO INFER THE
CHARACTERISTICS OF POPULATION.
ERROR: sampling error VS non-sampling error

SAMPLING METHODS
• 1. SIMPLE RANDOM SAMPLE: A SAMPLE SELECTED SO THAT EACH ITEM
IN THE POPULATION HAS THE SAME PROBABILITY OF BEING INCLUDED.
• 2. SYSTEMATIC RANDOM SAMPLE: A RANDOM STARTING POINT k IS
SELECTED, AND THEN EVERY kth MEMBER OF THE POPULATION IS
SELECTED.
• 3. STRATIFIED RANDOM SAMPLE : A POPULATION IS DIVIDED INTO SUB-
GROUPS, CALLED STRATA, AND A SAMPLE IS RANDOMLY SELECTED FROM
EACH STRATUM.
• 4.CLUSTER SAMPLING: A POPULATION IS DIVIDED INTO CLUSTERS USING
NATURALLY OCCURING GEOGRAPHIC OR OTHER BOUNDARIES. THEN,
CLUSTER ARE RANDOMLY SELECTED AND A SAMPLE IS COLLECTED BY
RANDOMLY SELECTING FROM EACH CLUSTER.
Sampling Methods
• Purposive random sampling
• Proportional random sampling
• Non-Random sampling
• Convenience sampling
• Judgment sampling
• Snowball sampling
HOW CAN SAMPLES BE USED TO MAKE ESTIMATES?
• A SAMPLE IS A TOOL TO INFER SOMETHING ABOUT
POPULATION OR STATISTICS CAN BE USED TO FIND
SOMETHING ABOUT A CHARACTERISTIC OF POPULATION OR
A PARAMETER.
EXAMPLE: X CAN BE USED TO ESTIMATE µ.
Central limit theorem :

If all samples of a particular size are selected from any
population, the sampling distribution of the sample mean is
approximately a normal distribution. This approximation
improves with larger samples.
SAMPLING DISTRIBUTION OF THE SAMPLE MEAN
• SAMPLING DISTRIBUTION OF THE SAMPLE MEAN IS A
PROBABILITY DISTRIBUTION OF ALL POSSIBLE SAMPLE MEANS OF
A GIVEN SAMPLE SIZE ( n ).
• THE FOLLOWING EXAMPLE ILLUSTRATES THE CONTRUCTION OF A
SAMPLING DISTRIBUTION OF SAMPLE MEAN.
• LET US ASSUME THAT A POPULATION CONSISTS OF FOUR
ELEMENTS X1 =1, X2 = 2, X3 = 3 AND X4 = 4. CONSIDER ALL THE
SAMPLES OF A GIVEN SIZE n = 2 THAT COULD BE DRAWN FROM
THIS POPULATION. (CONSIDER CAREFULLY WHETHER ONE IS
SAMPLING WITH OR WITHOUT REPLACEMENT).
Sampling with Replacement
All possible All possible
samples sample means samples sample means
1. 1.1 1.0 9. 3.1 2.0
2. 1.2 1.5 10. 3.2 2.5
3. 1.3 2.0 11. 3.3 3.0
4. 1.4 2.5 12. 3.4 3.5
5. 2.1 1.5 13. 4.1 2.5
6. 2.2 2.0 14. 4.2 3.0
7. 2.3 2.5 15. 4.3 3.5
8 2.4 3.0 16. 4.4 4.0
µ = ( 1 + 2 + 3 + 4 ) : 4 = 2.5 ( population mean )
µx = ( 1 + 1.5 + 2.0 +------------- + 4.0 ) : 16 = 2.5 The mean of all the means is
equal to the population mean or µ.
Variance and Standard Deviation
σ2= ( 1 – 2.5 )2+( 2 – 2.5 )2 +( 3 – 2.5 )2 + ( 4 – 2.5 )2 : 4
= ( 2.25 + 0.25 + 0.25 + 2.25): 4 = 5/4 or σ2 = 5/4 = 1.25
σ = √ 1.25 = 1.118 ( std deviation of population)
The standard deviation of the sample means is formally referred as
standard error of the mean.
σx2=( 1 -2.5 )2 + ( 1.5 – 2.5 )2 +------------------------- + ( 4 – 2.5 )2 : 16
= 10 : 16 = 0.625 and σx = √ 0.625 = 0.79 or
σ 1.118
σx = ----- = ----------- = 0.79
√n √2
Note that when we increase n, the σx will decrease.
Without Replacement
All possible samples Sample means
1.2 1.5
1.3 2.0
1.4 2.5
2.3 2.5
2.4 3.0
3.4 3.5
µx = ( 1.5 + 2.0 + ------- + 3.5 ) : 6 = 2.5
σx2 = ( 1.5 – 2.5 )2 + ( 2.0 – 2.5 )2 +----- + ( 3.5 – 2.5)2 : 6
= (1 + 0.25 + 0 + 0 + 0.25 + 1) = 2.5 : 6 = 0.417
σ
σx = √ 0.417 = 0.645 or σx = --- ( √ ( N – n ) : ( N – 1 )
√n
Example: Sampling Distribution of The Mean
• A population consists of three elements:
X1 = 25, X2= 30 and X3 = 35.
Calculate:
a. Sampling distribution of the mean if the sample
size = 2 ( with replacement and without
replacement ).
b. Variance and std deviation of the sample means.
1. Confidence interval for the population mean ( µ ) when n ≥ 30 ( Z
value )
• Confidence interval for µ is estimated by:
X ± a sampling error or X ± z value SE
Z value depends on confidence level.
SE = standard error of the mean
SE = σ/√n, if std deviation of population is known.
But, if the std deviation of population is unknown,
SE = s/√n , s = std deviation of sample.
Sampling Error and Confidence Interval Estimate
• Sampling Error : The diffrence between a sample

statistic and its corresponding population parameter.
This error will decrease as the sample size increase.
• Confidence interval estimate: A range of values
constructed from sample data so that the population
parameter is likely to within that range occur at
specified probability. The specified probability is called
the level of confidence.
EXAMPLE: Confidence Interval for µ ( n≥30 )
• The mean and std deviation for the quality grade point averages of a
random samples of 36 college seniors are calculated to be 2.9 and
0.3 respectively.
a. Find the 92 % confidence interval for the mean of the entire senior
class.
b. How large the sample size is required if we want to reduce the
sampling error not more than 0.05.
Confidence Interval for µ ( n < 30 )
• The weights of 7 similar boxes of cereal are 9.8, 10.2, 10.4,
9.8, 10.0, 10.2 and 9.6 ounces.
Find a 90% confidence interval for he mean of all such boxes of

cereal. ( use t – table ).
Step1: calculate the mean and std deviation of sample.
2: use t table to find critical value for 90% C.L ( tv).
3: substitute of all information from step 1 and 2
into the formula:
X - tv ( sd/√n) ≤ u ≤ X + tv ( sd/√n )
Level of Confidence
Required level value of α value of Z table
of confidence
90 % 10 % ± 1.65
95 % 5% ± 1.96
99 % 1% ± 2.58
Example : The prices at which certain type of instant coffee was
being sold on a given day were collected from a random sample
of 45 shops around the country. The mean price was $ 1.95 with
a standard deviation of $ 0.27. Compute a 80 % confidence
interval for the population mean.

Sampling Methods: - Reasons To Sample

Uploaded by

Copyright:

Available Formats

Sampling Methods: - Reasons To Sample

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sampling Methods: - Reasons To Sample

Uploaded by

Copyright:

Available Formats

Sampling Methods

ERROR: sampling error VS non-sampling error

Central limit theorem :

• Sampling Error : The diffrence between a sample

Find a 90% confidence interval for he mean of all such boxes of

You might also like