BASIC CONCEPTS IN

SAMPLING AND
SAMPLING TECHNIQUES
 SURVEY METHOD

•It is a method of collecting data by asking

people questions by either personal
interview, mailed questionnaire,
telephone, on-line survey, or focused
group discussion.
 Census vs Sample Survey

Census or Complete Enumeration

• the process of gathering information from every unit in
the population
Sample Survey
• the process of obtaining information from a subset unit
in the population
 Two Types of Sample Surveys

1. Descriptive
The main objective of a descriptive sample survey is simply to
obtain certain information about large population groups like number of
men and women who have finished college.

2. Analytical
Comparisons are made between subgroups of the population to
find out if there are differences that exist among them and to form or
verify hypotheses about the reasons for these differences.
Examples of Sample Surveys

 Quarterly Labor Force Survey

 Family Income and Expenditures Survey
 National Demographic and Health Survey
 Roles of Sampling in Statistical Work
1. Sampling in survey work
 The main purpose of conducting sample surveys is generally to
estimate certain population parameters

Examples: average income of families

average age of the population
participation rate of female workers in factories
most common occupation of employed persons
average production
fertility rate
infant mortality rate
 Information is obtained or desired from a sample
of statistical population, but inferences are required
on characteristics of the whole statistical population.
 A second possible purpose may be to test a
statistical hypothesis about the statistical population
Example: the hypothesis that at least 40% of the
households in Metro Manila have T.V. sets. To do this, a
sample of households in Metro Manila is selected and the
proportion of households possessing TV sets will be
calculated.
 Appropriate statistical test will be used to prove whether
the hypothesis is supported by the survey findings.
2. Use of sampling in the design and analysis of experiments
 In the conduct of experiments, whether indoor
experiments or outdoor experiments, different sampling
techniques are employed. Necessarily, the analysis of these
experiments would depend upon the sampling design. The
statistical population represents all possible applications of
several experimental “treatments” that can be used.
 Example: Most common applications of sampling in
designing experiments are in agriculture where different kinds of
fertilizers are being tested to find the rate of application that can
give the highest yield to a crop.
 Most of these experiments utilize experimental designs that
prescribe the analysis of observed or recorded information about the
experiment, usually what is called the analysis of variance or for
short ANOVA.

Example: Experiments in laboratories where a particular strain of

bacteria is being studied for resistance to antibiotics. The problem is
to design experiments so that the maximum amount of information
can be made available for inferences about the entire population,
estimated from a sample of limited size.
 Thus, an experiment is typically carried out in order to
measure the effect of some treatment.
3. Use of sampling in quality control work
 Quality control methods are usually used in industrial
establishments to test the quality of products coming out of a
machine. Inferences are needed on how well the products
conform to specifications.
Example: Sampling for quality control is to test the life span of
electric bulbs produced. The testing has to be done on a sample
basis to determine the average life span of an electric bulb. This
activity can only be done on a sample basis. Otherwise, the
company may not have any product to sell if done on a 100%
basis.
4. Broadening the scope of a census program
 In undertaking a census, be it population, housing,
agriculture and fisheries, the number of items that can be
covered with minimum cost is limited to very important
items which cannot be collected on a sample basis
because these are benchmark data that require complete
coverage.
 Among these items that are collected on a 100 per cent
basis are: population count, housing inventory and
number of holdings and number of fishing operators.
A convenient method of optimizing the cost of conducting a
census is the combination of complete enumeration and a
sample survey. The former method is used to gather
information on a number of basic items like what has been
mentioned earlier, and the latter to obtain additional items
that broadens the scope of the census. In other words, the
scope of the census is broadened by the use of sampling.
Example: In the Population Census, only basic demographic
characteristics of the population are taken on a 100 percent
basis while other socio-economic characteristics of the
population are obtained on a sample basis.
 Sources of Errors in Surveys
Two Major Groups of Survey Errors:
1. Errors of Non-observation
- where the sampled elements comprise only part of the target
population
- can be attributed to sampling, coverage, or nonresponse
2. Errors of Observation
- where recorded data deviate from truth
- can be attributed to interviewer (data collector) respondent,
instrument, or method of data collection
 Advantages of Sampling

1. Less expensive
 The use of sampling in the conduct of a survey is less
expensive compared with a complete coverage of the
population like a census when absolute precision is not
necessary.
 Also when funding is limited, it is still possible to collect
information through sampling with certain degree of
precision
2. Results are obtained faster with better quality of data
 Enumeration of a sample rather than a census
will definitely save time in data gathering as well as
processing of data aside from saving money.
 The duration of enumeration days will be
shortened and the volume of questionnaires to process
will be very much less so that results will come out
much earlier than the result from complete
enumeration.
Furthermore, there will be less enumerators to hire,
supervisors will have less to supervise which will redound to
better quality of data gathered.
Example: Release of the 1970 Census of Population and
Housing in a shorter span of time was made possible through
the use of sampling. A sample of 5% of census returns was
selected for advance processing and tabulation to generate
socio-economic characteristics of the population in less than
one year from the conduct of the census. Normally, it takes
more than 2 years to come out with the socio-economic
characteristics of the population from a census if all the
questionnaires will be processed.
3. Gather more detailed information
 When gathering very detailed data, it is not feasible
to consider them on a complete enumeration basis except
through a sample.
 The use of sampling permits particular attention to
be given to a limited number of cases.
Examples: family income and expenditures studies; health
and demographic studies on the population at large, such as
fertility, morbidity, mortality, mother and child care and
other health conditions and other social concerns studies.
4. Establish time series
 Use of sampling in the establishment of time
series data is the most practical and least expensive
way of obtaining data promptly.
Examples of time series data: Consumer Price
Index (CPI), Labor Force statistics, and Family
Income and Expenditures statistics including poverty
statistics.
 The frequency of the series depends upon the type of
survey conducted.
Examples: The survey that gathers data used as inputs to
the computation of the CPI is done every month; the labor
force survey is conducted every quarter; the Family Income
and Expenditures Survey is conducted every 3 years; the
National Health and Demographic Survey is usually
conducted every 3 years; and the Crop and Livestock
Survey is also conducted every quarter.
5. Use of sampling to control non-sampling errors
 A survey is a chain of operations that extends from
data collection to publication of survey results. Varying
degrees of control can be exerted on the quality of each
link in the chain. This is also true for a census operation. It
is much easier to control non-sampling errors that creep in
the various stages of operation in surveys because its
magnitude is less than in a census operation.
 It is, therefore, accepted that sampling is employed
to control non-sampling errors.
 Definition of Terms
Example: In a certain community, an opinion poll was
conducted to determine public sentiment toward a bond
issue in an upcoming election. The objective of the survey
was to estimate the proportion of voters in the community
who favored the bond issue.
1. Element
- is an object on which a measurement is taken.
Example: In the given example, an element is a registered
voter in the community.
 Definition of Terms
2. Target Population
- the population about which information is desired.
Example: The target population in the given
example is the collection of voters in the
community.
3. Sampled Population
- the population from which a sample is actually
taken
 Definition of Terms
Notes about target and sampled population:
a) Ideally, the target and the sampled populations should
be the same collection of elements.
b) The target population is the collection under study, that
is, we wish to infer on the characteristics of this
population. However, if we cannot get a sample from this
population then get another similar population, which we
call the sampled population, from where we obtain the
sample.
 Definition of Terms
Notes about target and sampled population:
c) If we take the sample from the sampled population,
then, whatever generalizations we make using the sample
will merely apply to the sampled population.
d) If the sampled population coincides with the target
population then we can extend the generalizations on the
sampled population to the target population, which is the
true population of interest.
 Example of Target and
Sampled
The Populations
investment expenditures are important to the businesses in a
province containing a university. How can we estimate the average
monthly amount spent on entertainment per student? To locate students
who may be randomly selected from a directory is difficult, but to locate
randomly selected classrooms, all of which should contain students at a
prime class hour, like 10:00 A.M. on Monday, is relatively easy. Since
classes may be large, sufficient information can be obtained by sampling
a subset of those students in each sampled class.
A certain mid-sized university has 12,000 students divided into 150
classes at 10:00 A.M. on Mondays. Almost all the students should be in
class at this hour. Four classes are randomly selected and approximately
10% of the students in each class are interviewed.
What are the target and sampled populations in this case?
 Example of Target and
Sampled
Populations

Target Population: All 12,000 students at the

university.
Sampled Population: Students interviewed from
the four randomly selected classes (10% of the
students in those classes).
 Example of Target and
Sampled
Populations
4. Sampling Units
- are nonoverlapping collections of elements
from the population that cover the entire
population
Example: In the bond issue example, a
sampling unit may be a registered voter in the
community.
 Example of Target and
Sampled
6. Sample Populations
- is a collection of sampling units drawn from a frame or
frames.
Example: Sampling unit - individual voter
Frame - list of registered voters
Sample - number of voters in the public opinion poll
The sampling units will be contacted to determine their
preference for the upcoming bond issue. We then use the
information obtained from these voters to make an inference
about voter preference throughout the community.
 Example of Target and Sampled
Populations
7. Probability of Selection
- The probability of selection is the chance that each unit in
the population has of being included in the sample. Probability
values range from 0 to 1. Usually the probability of selection is
determined from the elements of the sampling frame.
8. Statistic
- A statistic is an estimate computed from sample
observations of a characteristic, usually for the purpose of
making an inference about the statistical population.
Two Methods of Sampling
 Two Methods of Sampling

1. Probability Sampling
• procedure wherein every element of the
population is given a (known) nonzero chance
of being selected in the sample
 Two Methods of Sampling

2. Nonprobability Sampling
• procedure wherein not all the
elements in the population are given a
chance of being included in the sample
 Methods of Nonprobability
Sampling

•Probabilities of selection are not

specified for the individual units of
the population in nonprobability
sampling.
 Methods of Nonprobability
Sampling
• The clear disadvantage of nonprobability
sampling is that the researcher cannot assert
that the sample is representative of the larger
population because the probability that a
person or object will be chosen is not known.
 Methods of Nonprobability
Sampling
• The clear disadvantage of nonprobability
sampling is that the researcher cannot assert
that the sample is representative of the larger
population because the probability that a
person or object will be chosen is not known.
 Methods of Nonprobability
Sampling
•The advantages of nonprobability
sampling are convenience in terms of
getting subjects, less costly and much less
difficult than methods of probability
sampling.
 Methods of Nonprobability
Sampling
•A nonprobability sample is desirable if the
researcher has no intention of generalizing
the conclusions beyond the sample and if
the study is only a test run for a bigger
study.
 Methods of Nonprobability
Sampling
Purposive Sampling
• sets out to make the sample agree with the
population in regard to certain characteristics
Quota Sampling
• a specific number of particular types of
elements are selected
 Methods of Nonprobability
Sampling
Convenience Sampling
•chooses units which come to hand or
are convenient
Judgment Sampling
•selects sample in accordance with an
expert’s judgment
 Cases Wherein
Nonprobability Sampling is Useful

Only few are willing to be interviewed

Extreme difficulties in locating or
identifying subjects
Probability sampling is more
expensive to implement
Methods of Probability Sampling

• Probability Sampling is a method of

selecting a sample wherein each unit in
the population (for example, a person) has
a known non-zero probability of being
included in the sample. This probability is
used in computing estimates of population
characteristics.
Methods of Probability Sampling

• It specifies rules and procedures for both

sample selection and estimation. Used if
the main objective of the sample survey is
making inferences about the
characteristics of the population under
study.
 Probability vs. Non-Probability Sampling
ISSUE: Which is “better”?
 There is no straightforward answer to resolve this issue
outright.
 Non-probability sampling may be “cheaper” to implement.
 It is only with probability sampling that one is able to derive
objective measures of “errors”.
 The strength of one method is the weakness of the other.
However, the cost of doing a survey can be minimized. With
this, the use of probability sampling is recommended.
 Probability vs. Non-Probability Sampling
ISSUE: Which is “better”?
 There is no straightforward answer to resolve this issue
outright.
 Non-probability sampling may be “cheaper” to implement.
 It is only with probability sampling that one is able to derive
objective measures of “errors”.
 The strength of one method is the weakness of the other.
However, the cost of doing a survey can be minimized. With
this, the use of probability sampling is recommended.
Methods of Probability Sampling

* Simple Random Sampling

* Stratified Sampling
* Systematic Sampling
* Cluster Sampling
* Multi-Stage Sampling
 Simple Random Sampling (SRS)

Description of the Design

• Random sampling is the process of selecting a
sample of size (n) giving each sampling unit an equal
chance of being included in the sample. An SRS of n
observations of the population is a sample that is
chosen in such a way that each subset of n
observations of the population has the same chance
of being selected.
 Simple Random Sampling (SRS)

• Random sampling may be with

replacement (SRSWR) or Without
replacement (SRSWOR). In SRSWR, a
chosen element is always replaced before
the next selection is made, so that an
element may be chosen more than once.
 Sample Selection Procedure

1. Make a list of the sampling units and number them

from 1 to N, where N is the population size.
2. Select n (distinct for SRSWOR, not necessarily
distinct for SRSWR) random numbers from 1 to N
using some random process; for example, the table of
random numbers or lottery. The sample consists of the
units corresponding to the selected random numbers.
 barangays in Mandaluyong City. Use the table of random
numbers (see Appendix 1) to determine which of the barangays
Example
listed oftoSimple
below are Random
be included Sampling
in the sample.
• Addition Hills Hagdang Bato
Itaas New Zaniga
• Bagong Silang Hagdang Bato
Libis Old Zaniga
• Barangka Drive HarapinAng Bukas
Pag-asa
• Barangka Ibaba Highway Hills
Plainview
• Barangka Ilaya Hulo
Pleasant Hills
• Barangka Itaas Mabini-J.Rizal
Poblacion
• Buayang Bato Malamig
San Jose
Advantages
• The theory involved is much easier to
understand than the theory behind other
sampling designs.
• Estimation methods are simple and easy.
Disadvantages
• The sample chosen may be widely spread,
thus entailing high transportation costs.
• A population list, or frame, is needed.
• The sample chosen may not be truly typical
of the population if the population is
heterogenous with respect to the characteristic
 Stratified Sampling
Description of the Design
There are cases wherein the population consists of item which
are heterogeneous with respect to the characteristic under
study. In such situations the population should be divided, or
stratified, into more or less homogeneous subpopulations or
strata before sampling is done.
Stratified random sampling then consists of selecting an SRS
from each of the strata into which the population has been
divided.
 Sample Selection Procedure

1. Stratify the population into strata so that

ideally each stratum will consist of more or
less homogeneous units.
2. After the population has been stratified,
an SRS is selected from each stratum.
Illustration
 Stratified Sampling
Frequently, the number of units selected from each
stratum is proportional to the number of units in that
stratum in the population.
Stratified sampling cannot be used unless some
information concerning the population and its strata is
available.
An extremely important point, which is often
overlooked, is that the strata must be ones, which are
related to the topic being studied.
 Example of Stratified Sampling

• An advertising firm, interested in how much to emphasize television

advertising in a certain town, decides to conduct a sample survey to
estimate the average number of hours per week that households
within the town watch television. The town contains two municipalities,
Municipality A and Municipality B, and a rural area. Municipality A is
built around a factory, and most households contain factory workers
with school-aged children. Municipality B contains older residents with
few children at home. There are 155 household in Municipality A, 62 in
Municipality B, and 93 in the rural area. Discuss the merits of using
stratified random sampling in this situation.
• Stratification may bring about a gain in
precision of the estimates of characteristics of
the population.
• It allows for more comprehensive data
analysis since information is provided for each
stratum.
• It is administratively convenient.
Disadvantages
• A listing of the population for each stratum is
needed.
• The stratification of the population may
require additional prior information about the
 Systematic Sampling

Description of the Design

• Systematic sampling with a random start is a method of selecting
a sample by taking every kth unit from an ordered population, the
first unit being selected at random. Here k is called the sampling
interval and 1/k the sampling fraction.
Sample Selection Procedure
1. Number the units of the population consecutively from 1 to N.
2. Determine k by the formula k = N / n
= population size/sample size
 Systematic Sampling

3. Select the random start r (between 1 and k). The unit

corresponding to r is the first unit of the sample.
4. The other units of the sample correspond to r + k, r + 2k, ...
Example of Systematic Sampling
A medical investigator is interested in obtaining information
about the average number of times N = 15,000 specialists
prescribed a certain drug in the previous year. To obtain a
sample of n = 1600 specialists, we could select one specialist at
random from the first k = 9 names appearing on the list and then
select every ninth name thereafter until a sample of size 1600 is
selected.
Illustration
 Example of Systematic Sampling

Suppose that it is desired to estimate the average

number of owner-occupied dwelling units per block in
a given area of N = 24 blocks. If there is a map of the
area with the blocks distinctly identified, a systematic
sample from the map can easily be chosen.
Sample-Selection Procedure:
Frame: Map of the area with blocks distinctly identified, or
a list of blocks in the area arranged in order of location
Sampling Unit: Block
Suppose that it is desired to choose 6 blocks out of 24. k = 4
Example: From a population of 24 items, choose a sample
of size 6. Random start r = 15, K = 4
Method A:
1. Number the blocks consecutively from1 to 24 in the frame.
2. Select a number at random between 1 and 4.
3. If the selected random number is 3, then choose every 4th unit
thereafter and these are 7, 11, 15, 19, and 23.
Method B:
4. Number the blocks consecutively from 1 to 24 in the frame.
5. Select a number at random from 1 to 24. Suppose the selected
number is 15.
6. The block numbered “15” is in the sample. From the block
numbered “15”, take every fourth block, that is 19, 23, 3, 7, and
11.
Advantages
 Drawing of the sample is administratively easy.
 It is possible to select a sample in the field without
a frame.
Disadvantage
 If periodic regularities are found in the list, a
systematic sample may consist only of similar types.