DR. ROGER L.

NUQUI
Sampling is a way of getting a
representative portion of the target
population.

Regardless of the research design

(descriptive or experimental) used by a
researcher, sampling is needed for
practical reasons especially when the
population is very large. A total
population of 100 or less may not need
to be sampled but more than this,
sampling is desirable for effectiveness,
efficiency and economy in data
gathering.
Population and Sample
Population and sample are two
important concepts in research
particularly in determining the
sources of data. Population refers
to the total number of people a
researcher wants to study. Sample
refers to the people chosen to
represent the population.
Sampling is needed in a research
involving a large population that
may be limited by human
resources, money, materials and
machinery. According to Calmorin
and Calmorin (2012), the use of
total population is advisable for a
population of a hundred or less.
More than a hundred requires
sampling. The strengths and
weaknesses of sampling are
enumerated by Calmorin and
Calmorin (2012).
Strengths:
Time, money, and effort are minimized. By sampling, the number of
respondents, subjects or items to be studied becomes small yet
they represent the population. As such, data collection, analysis
and interpretation are lessened.

Sampling is more effective. Since every individual in the population is

given a chance to be selected through sampling, data are
scientifically gathered, analyzed and interpreted.

Research is made faster and cheaper. By selecting a small portion of

the population as representative, collection, analysis, and
interpretation of data are faster and cheaper.

Sampling makes research more accurate. Because of the small size of

the data collected from a small number of sources, collection,
tabulation, presentation, analysis, and interpretation have fewer
errors as compared to voluminous data from the whole
population.

Sampling gives more comprehensive information. With a small sample

representing a big population, a thoroughly investigated study
can yield results that give more comprehensive information that
allows generalization and conclusions.
Weaknesses:
Due to the limited number of data source, detailed sub-
classification must be prepared with utmost care.

Incorrect sampling design or incorrectly following the

sampling plan will obtain results that are misleading.

Sampling requires an expert to conduct the study in an

area, otherwise the results obtained will be
erroneous.

The characteristic to be observed may occur rarely in a

population. For example, teacher over 30 years of
teaching experience.

Complicated sampling plans are laborious to prepare.

Two General Types of Sampling:

Probability sampling or scientific

sampling gives every member of the
population equal chance to be selected
as part of the study.

Five Sampling Schemes of this Type:

– Simple random sampling

– Systematic sampling
– Stratified sampling
– Cluster sampling
– Multi-stage sampling
Simple random sampling is
made when all the members
of the population are given a
chance to be selected.
Selection is done by draw lot
or the use of the table of
random numbers.
Simple random sampling

• Get a list of the total population.

• Cut pieces of paper to small sizes (1 x 1
in.) that can be rolled.
• Put a number on each piece of paper
corresponding to each number of the
total population
• Roll each piece of numbered paper and
put them in a box.
• Shake well the box to give equal chance
for every number to be chosen as sample.
• Pick out one rolled paper at a time and
unroll it.
• Record the number of the unrolled paper.
• Repeat picking out until the desired
number of sample is completed.
Systematic sampling is choosing the nth
name in a population as the sample. It
entails using a list of the population and
deciding how the nth name is chosen. For
example, the population of students in one
school is listed alphabetically and
numbered consecutively. From the list, the
sample to be taken is the name that falls
every nth in the list until the desired
number of sample is completed. So with a
population of 500 and 250 as the desired
sample, the individual sample may be the
name that falls in every count of two or all
those that are in the odd number or the
even number.
Systematic Sampling

This sampling scheme is used when there is a ready list

of the total population.
The steps in using this schemes are:
• Get the list of the total population.
• Divide the total population by the desired sample
size to get the sampling interval.

Example:
If the total population is 5000 and the desired sample
is 100,

Sampling Interval = 5000 = 50

100
Get the No. 50 as the first sample and every 50th person
in the list or 100, 150, etc. until 100 respondents are
completed.
Stratified random sampling is
dividing the population into strata
and drawing the sample at random
from each division. For example,
the population to be sampled are
the elementary school teachers in
one district. These are stratified by
their ranks as teacher 1, 2, 3.
Samples of equal number are
drawn from each rank
Stratified Sampling

This is used when there is a ready list of the population whose

members are categorized as students, farmers, fishermen.
The steps to follow are:

Get the list of the total population.

Decide on the sampling size or the actual percentage of the
population to be considered as sample.
Get an equal proportion of sample from each group.

Example:

500 pupils x .20 = 100

200 teachers x .20 = 40
150 parents x .20 = 30
Total Sample 170

Get the 170 respondents by simple random or systematic

sampling.
Cluster sampling is a design that
uses a group as sample rather than
an individual. For example, the
population may be the parents in
one school district. The parents
may be grouped by barangay
within the district or by those in
the east, west, north and south of
the district. From these groupings,
the sample cluster is chosen
randomly or systematically. This
differs from stratified sampling that
includes all the strata in the
sampling process.
Cluster Sampling
This is used when the population is
homogeneous but scattered geographically
in all parts of the country and that there is
no need to include all in the sampling.

Following are the steps in this scheme:

Decide on the sample size. Select the

geographical area that will serve as the
cluster.

Select the sample within this area/cluster by

systematic or stratified sampling depending
on the availability of the information.
Multi-stage sampling is done by stages:
two, three, four as the case may be
depending on the number of stages
sampling is made. Here the population
is grouped by hierarchy from which
sampling is done in each stage. For
example, the population to be studied
consists of the personnel in the public
elementary schools in the country. So
samples have to be taken from the
national, regional, provincial, district,
and school levels
Multi-stage sampling

As the name suggests, sampling is done by stages or at

each level. The steps followed in this scheme are:

Decide on the level of analysis to be studied such

national, regional, provincial, district, barangay, etc.

Select the sample at the next lower level. For example if

the survey is provincial, the survey should start at the
municipal level, then district, then barangay, etc.
Example:

1st level, 3 municipalities/province

2nd level, 2 districts/municipality
3rd level, 4 barangays/districts
4th level, 100 respondents/barangay

Use any of the sampling techniques given earlier in

arriving at the desired sample number.
2. Non-probability or non-scientific
sampling is the opposite of
probability. Here the researcher’s
judgment determines the choice of the
sample.

Three kinds commonly used:

Purposive Sampling- In this design,

the samples are chosen based on the
judgment of the researcher who
determines an individual as sample for
possessing special characteristics of
some sort.
• Incidental/Accidental Sampling –
As the term implies, this design is
used to take samples who may be
the most available or the nearest at
the time of data gathering.

• Quota Sampling- A design

popular for opinion research, this
sampling is made by looking for
individuals that possess the
required characteristics or
prescribed criteria of the research.
Steps in Sample Selection

Each of the above sampling schemes follows certain

procedures in determining the sample. Mercado (1985)
suggested these steps, to wit:

Incidental/Accidental

Decide the sample size.

Look for people who can be interviewed.
If the sample decided in 100, interview the first 100 people
you meet.

Purposive sampling

Decide on the sample size by looking at the table of sample

size in statistics books.
Decide on the criteria in choosing the sample. These criteria
may include age, sex, and education.
Choose the sample based on the criteria. For example if the
criterion is age group 21 to 30 years old, then only
respondents belonging to this age group are interviewed.
Determining the Sample Size

For a population of more than 100, sampling is a must. Following is a

formula for scientific sampling, illustrations and examples
(Calmorin and Calmorin, 1995):

• Ss = NV + [Se (1 - p)]____
Nse + [V x p(1 - p)]
Where Ss = Sample size
N = Total number of population
V = The standard value (2.58) of 1 percent level of
probability with
0.99 reliability
Se = Sampling error
P = The largest possible proportion (0.50)

To illustrate the above formula, these steps have to be followed:

1. Determine the total population (N) assumed to be studied.
2. Get the value of V (2.58), Se (.01) and p (0.50).
3. Compute the sample size using the above formula
The total population is 500 has a
standard value of 2.58 at 1 percent
level of probability and 99 percent
reliability. The sampling error is 1
pecent (0.01) and the proportion of
a target population is 50 percent
(0.50).
The sample size is computed as follows:

N = 500
V = 2.58
Se = 0.01
P = 0.50
Ss = NV + (Se)2 x (1-P)
Nse + (V)2 x P(1-P)

= 500(2.58) + (0.01)2 x (1-0.50)

500(0.01) + (2.58)2 x 0.50(1-0.50)

= 1290 + 0.0001 x 0.50

5 + 6.6564 x 0.50(0.50)

= 1290 + 0.00005
6.6641

= 1290.00005
6.6641

Ss = 193.57 or 194

This sample size of 194 represents the 500 subjects of

study.