2. INTRODUCTIONâĻ
ī Sampling is a process of selecting representative
units from an entire population of a study.
ī Sample is not always possible to study an entire
population; therefore, the researcher draws a
representative part of a population through
sampling process.
ī In other words, sampling is the selection of some
part of an aggregate or a whole on the basis of
which judgments or inferences about the
aggregate or mass is made.
ī It is a process of obtaining information regarding a
phenomenon about entire population by examining3/2/20152
4. īPopulation:
ī Population is the aggregation of all the
units in which a researcher is interested.
In other words, population is the set of
people or entire to which the results of a
research are to be generalized.
ī For example, a researcher needs to
study the problems faced by
postgraduate nurses of India; in this the
âpopulationâ will be all the postgraduate
nurses who are Indian citizen. 3/2/20154
5. Population is defined as The Entire Group under study. Sometimes it is also called as
the âUniverse.â
Population
6. Target Population:
3/2/20156
ī A target population consist of the total number of
people or objects which are meeting the
designated set of criteria.
ī In other words, it is the aggregate of all the
cases with a certain phenomenon about which
the researcher would like to make a
generalization.
ī For example, a researcher is interested in
identifying the complication of diabetes mellitus
type-II among people who have migrated to
Canada. In this instance, the target population
are all the migrants at Canada suffering with
7. Accessible population:
3/2/20157
ī It is the aggregate of cases that conform to
designated criteria & are also accessible as
subjects for a study.
ī For example, âa researcher is conducting a study
on the registered nurses (RN) working in Father
Muller Hospital, Mangalore â. In this case, the
population for this study is all the RNs working in
Father Muller Hospital, but some of them may be
on leave & may not be accessible for research
study. Therefore, accessible population for this
study will be RNs who meet the designated
criteria & who are also available for the research
9. Sample:
3/2/20159
ī Sample may be defined as
representative unit of a target
population, which is to be worked
upon by researchers during their
study.
ī In other words, sample consists of a
subset of units which comprise the
population selected by investigators
or researchers to participates in their
10. Element:
3/2/201510
ī The individual entities that comprise the
samples & population are known as
elements, & an element is the most basic
unit about whom/which information is
collected. An elements is also known as
subject in research. The most common
element in nursing research is an
individual. The sample or population
depends on phenomenon under study
11. Sampling frame:
3/2/201511
ī It is a list of all the elements or subjects
in the population from which the sample
is drawn.
ī Sampling frame could be prepared by
the researcher or an existing frame
may be used.
ī For example, a researcher may prepare
a list of the all the households of a
locality which have pregnant women or
may used a register of pregnant women
12. Sampling error:
3/2/201512
īThere may be fluctuation in the
values of the statistics of
characteristics from one sample to
another, or even those drawn from
the same population.
15. Sample size
3/2/201515
Size of sample should be determined by a researcher
keeping in view:
1. Nature of universe: -homo (small sample)
- hetero (large sample).
ī 2. No. of classes proposed: directly proportional to
the sample size .
ī 3. Nature of study: general (large)
ī intensive (small).
ī 4.Type of sampling: small random sample is better
than a large but bad one.
16. ContdâĻ
3/2/201516
Standard of accuracy: high level of precision large
sample.
6. Availability of finance: sample size =amount of
money available.
ī 7. Other considerations: size of population,
ī size of questionnaire,
ī nature of units,
ī conditions.
21. ī Economical: In most cases, it is not possible &
economical for researchers to study an entire
population. With the help of sampling, the researcher
can save lots of time, money, & resources to study a
phenomenon.
ī Improved quality of data: It is a proven fact that when
a person handles less amount the work of fewer
number of people, then it is easier to ensure the quality
of the outcome.
ī Quick study results: Studying an entire population
itself will take a lot of time, & generating research
results of a large mass will be almost impossible as
most research studies have time limits
ī Precision and accuracy of data: Conducting a study
on an entire population provides researchers with
voluminous data, & maintaining precision of that data
3/2/201521
22. CHARACTERISTICS OF GOOD SAMPLE
īTrue Representative
īFree from sample bias and errors
īNo substitution and incompleteness
īAppropriate sample size(Optimum
size (adequately large)
3/2/201522
23. ContdâĻ
īHas all characteristics that are present
in population
īEconomically viable
īResults can be applied to the universe
in general with a reasonable level of
confidence or reliability
24. SAMPLING PROCESS
Identifying and defining the target
population
Describing the accessible population &
ensuring sampling frame
Specifying the sampling unit
Specifying sampling selection methods
3/2/201524
26. Sampling Design Process
Define Population
Determine Sampling Frame
Determine Sampling Procedure
Probability Sampling
Simple Random Sampling
Stratified Sampling
Cluster Sampling
Systematic Sampling
Multistage Sampling
Non-Probability Sampling
Convenient
Judgmental
Quota
Snow ball Sampling
Determine Appropriate
Sample Size
Execute Sampling
Design
27. Steps in Sampling Process
1. Define the population
2. Identify the sampling frame
3. Select a sampling design or
procedure
4. Determine the sample size
5. Draw the sample
32. 32
METHODS OF SAMPLING
īąThe methods of sampling can be divided
on the basis of the element of probability
associated with the sampling technique.
Probability means chances available to
members of the population for getting
selected in the sample. Accordingly, the
methods of sampling are classified into two
broad types:
ī Probability Sampling
īNon Probability Sampling
34. Classification of Sampling Methods
Sampling
Methods
Probability
Samples
Simple
Random
Cluster
Systematic Stratified
Non-
probability
QuotaJudgment
Convenience Snowball
Multistage
36. 36
Probability Sampling Method
ī Probability Sampling is also known as Random
Sampling or formal sampling
ī Probability means chance
ī Therefore element of the population has the equal
chance or opportunity of being selected in the
sample
ī Probability samples are more accurate
ī Eg. If a sample of 100 students is to be selected
from a population of 1000 students, then it is
known to every one that each student has 1000 /
37. 3/2/201537
ConceptâĻ
ī It is based on the theory of probability.
ī It involve random selection of the
elements/members of the population.
ī In this, every subject in a population has
equal chance to be selected sampling
for a study.
ī In probability sampling techniques, the
chances of systematic bias is relatively
less because subjects are randomly
38. Features of the Probability
sampling
3/2/201538
ī It is the only systematic and objective method of sampling that
provides equal chance to every element of the population in
getting selected in the sample
ī The results of probability sampling more accurate and reliable
ī It helps in the formulation of a true representative sample by
eliminating human biases
ī Under probability method each element of population known
in advance about the possibility of being included in the sample
ī The advantage of using a random sample is the absence of both
systematic & sampling bias.
ī The effect of this is a minimal or absent systematic bias, which is a
difference between the results from the sample & those from the
population.
40. Types of the probability sampling
3/2/201540
1.Simple random sampling
2.Stratified random sampling
3.Systematic random sampling
4.Cluster/multistage sampling
5.Sequential sampling
42. DR G K KALKOTI42
Simple Random Sampling
īą It is the basic probability sampling
technique and all other methods are
variations of simple random method.
īą It can be defined as the method of sampling
which provides every element in the
population an equal and known chance of
being selected in the sample.
ī Simple random can be done by
A) Lottery Method
B) Random Tables
c)The use of computer
43. Simple random sampling
3/2/201543
ī The entire process of sampling is done
in a single step, with each subject
selected independently of the other
members of the population
ī There is need of two essential
prerequisites to implement the simple
random technique:
1.Population must be homogeneous &
2.Researcher must have list of the
elements/members of the accessible
45. Steps
3/2/201545
ī The first step of the simple random sampling
technique is to
ī identify the accessible population & prepare
a list of all the elements/members of the
population.
ī The list of the subjects in population is
called as sampling frame & sample drawn
from sampling frame by using following
methods:
īąThe lottery method
īąThe use of table of random numbers
īąThe use of computer
46. 3/2/201546
1.The lottery methodâĻ
ī It is most primitive & mechanical method.
ī Each member of the population is assigned
a unique number.
ī Each number is placed in a bowel or hat &
mixed thoroughly.
ī The blind-folded researcher then picks
numbered tags from the hat.
ī All the individuals bearing the numbers
picked by the researcher are the subjects
for the study.
47. 3/2/201547
The use of table of random
numbersâĻ
ī This is most commonly & accurately used method
in simple random sampling.
ī It provides use of random numbers specially
designed for sampling purposes
ī Random table present several numbers in rows &
columns.
ī Researcher initially prepare a numbered list of the
members of the population, & then with a blindfold
chooses a number from the random table.
ī The same procedure is continued until the desired
number of the subject is achieved.
ī If repeatedly similar numbers are encountered, they
are ignored & next numbers are considered until
desired numbers of the subject are achieved.
ī Such type of random table are mostly found at the
end of statistical textbooks
49. 3/2/201549
The use of computerâĻ
ī Nowadays random tables may be
generated from the computer , & subjects
may be selected as described in the use of
random table.
ī For populations with a small number of
members, it is advisable to use the first
method, but if the population has many
members, a computer-aided random
selection is preferred.
51. Types of Simple Random Sample
īļWith replacement
īļWithout replacement
52. īļWith replacement
The unit once selected has the chance for again
selection
īļWithout replacement
The unit once selected can not be selected
again
53. REPLACEMENT OF SELECTED
UNITS
53
ī Sampling schemes may be without replacement
('WOR' - no element can be selected more than once
in the same sample) or with replacement ('WR' -
an element may appear multiple times in the one
sample).
ī For example, if we catch fish, measure them, and
immediately return them to the water before
continuing with the sample, this is a WR design,
because we might end up catching and measuring
the same fish more than once. However, if we do not
return the fish to the water (e.g. if we eat the fish), this
54. Merits and Demerits
3/2/201554
Merits
ī Ease of assembling
the sample
ī Fair way of selecting a
sample
ī Require minimum
knowledge about the
population in advance
ī It unbiased probability
method
ī Free from sampling
errors
ī Demerits
ī It requirement of a
complete & up-to-date list
of all the members of the
population.
ī Does not make use of
knowledge about a
population which
researchers may already
have.
ī Lots of procedure need to
be done before sampling
ī Expensive & time-
consuming
56. STRATIFIED SAMPLING
īļPopulation is divided on the basis of
characteristic of interest in the population e.g.
male and female may have different
consumption patterns.
57. Stratified Random Sampling
3/2/201557
ī This method is used for heterogeneous
population.
ī It is a probability sampling technique wherein the
researcher divides the entire population into
different homogeneous subgroups or strata, &
then randomly selects the final subjects
proportionally from the different strata.
ī The strata are divided according selected traits of
the population such as age, gender, religion,
socio-economic status, diagnosis, education,
geographical region, type of institution, type of
care, type of registered nurses, nursing area
specialization, site of care, etc.
58. DR G K KALKOTI58
Stratified Random Sampling
ī In this method, the population is divided and
subdivided with homogeneous or similar
characteristics
ī For example, a group of 200 college teachers can be
first divided into teachers in Arts faculty, Commerce
Faculty and Science Faculty.
ī After dividing the entire population of teachers into
such classes called strata, a sample is selected from
each stratum of teachers at random. These samples
are put together to form a single sample.
ī Stratified random sampling is more accurate and
representative as compared to simple random
sampling because under this the population is divided
into homogeneous groups.
59. Stratified Random Sampling
Population is divided into two or more groups
called strata, according to some criterion, such as
geographic location, grade level, age, or income,
and subsamples are randomly selected from each
strata.
Elements within each strata are homogeneous, but
are heterogeneous across strata
61. Types of Stratified Random Sampling
īļProportionate Stratified Random Sampling
Equal proportion of sample unit are selected from each
strata
īļDisproportionate Stratified Random Sampling
Also called as equal allocation technique and sample unit
decided according to analytical consideration
65. SYSTEMATIC SAMPLING
If a sample size of n is desired from a population
containing N elements, we might sample one
element for every n/N elements in the population.
66. Systematic Random Sampling
3/2/201566
ī It can be likened to an arithmetic progression,
wherein the difference between any two
consecutive numbers is the same.
ī It involves the selection of every Kth case from list
of group, such as every 10th person on a patient list
or every 100th person from a phone directory.
ī Systematic sampling is sometimes used to sample
every Kth person entering a bookstore, or passing
down the street or leaving a hospital & so forth
ī Systematic sampling can be applied so that an
essentially random sample is drawn.
67. CountâĻ
3/2/201567
ī If we had a list of subjects or sampling frame, the
following procedure could be adopted. The desired
sample size is established at some number (n) &
the size of population must know or estimated (N).
Number of subjects in target
population (N)
K = N/n or K=
Size of sample
ī For example, a researcher wants to choose about
100 subjects from a total target population of 500
people. Therefore, 500/100=5. Therefore, every 5th
person will be selected.
68. DR G K KALKOTI68
Systematic Sampling
ī It is modification of simple random
sampling. It is called as quasi-random
sampling.
ī It is called quasi because it is in between
probability and non-probability sampling.
ī The procedure of quasi sampling begins
with finding out the sample interval. This
can be found out by the ratio of the
population to the sample. Afterwards a
random number is selected from the
sample interval.
69. DR G K KALKOTI69
Illustration of Systematic
Sampling
ī Selecting a sample of 100 students out of 1000, the
sample interval will be 1000 divided by 100 i.e.10.
ī Then make small chits bearing numbers 1to 10 and put
them into a box
ī Then by using lottery method withdraw one slip and
suppose we get number 5 then proceed to select numbers
starting with 5 with a regular interval of 10.
ī The selected sample consists of elements bearing nos.
5,15,25,..........105,115 and so on .
ī It should be noted that up to selecting no.5,Systematic
sampling can be treated as probability sampling and
afterwards it is non-probability because the chances of
other elements are certainly affected
ī In this example numbers other than 5 have no chance of
being selected
70. Systematic Random Sampling
Order all units in the sampling frame based on some
variable and then every nth number on the list is
selected
Gaps between elements are equal and Constantī
There is periodicity.
N= Sampling Interval
71. Systematic Sampling
ī Order all units in the sampling frame
based on some variable and number
them from 1 to N
ī Choose a random starting place from 1
to N and then sample every k units after
that
3. Selecting a Sampling
Design
72. systematic random sample
number the units in the population from 1 to
N
decide on the n (sample size) that you want
or need
k = N/n = the interval size
randomly select an integer between
1 to k
then take
every kth unit
76. Cluster or Area Random Sampling
Clusters of population
units are selected at
random by dividing the
population into clusters
(usually along geographic
boundaries) and then all
or some randomly chosen
units in the selected
clusters are studied.
77. Cluster or multistage Sampling
3/2/201577
ī It is done when simple random sampling is almost
impossible because of the size of the population.
ī Cluster sampling means random selection of sampling
unit consisting of population elements.
ī Then from each selected sampling unit, a sample of
population elements is drawn by either simple random
selection or stratified random sampling.
ī This method is used in cases where the population
elements are scattered over a wide area, & it is
impossible to obtain a list of all the elements.
ī The important thing to remember about this sampling
technique is to give all the clusters equal chances of
being selected.
78. ī Geographical units are the most commonly used
ones in research. For example, a researcher wants
to survey academic performance of high school
students in India.
īŧHe can divide the entire population (of India) into
different clusters (cities).
Then the researcher selects a number of clusters
depending on his research through simple or
systematic random sampling.
īŧThen, from the selected clusters (random selected
cities), the researcher can either include all the high
school students as subjects or he can select a
number of subjects from each cluster through
CountâĻ
3/2/201578
79. DR G K KALKOTI79
Cluster Sampling
ī Cluster means group, therefore, sampling
units are selected in groups.
ī Cluster sampling is an improvement over
stratified sampling. Both simple random and
stratified random sampling are not suitable
while dealing with large and geographically
scattered populations. Therefore, large-scale
sample surveys are conducted on cluster
sampling basis.
ī The working of cluster sampling is based on
the principle that it is beneficial to use a large
sample of units closer to each other than to
select a small group of sample scattered over
a wider area.
80. DR G K KALKOTI80
Illustration of Cluster Sampling
ī Suppose researcher wants to study the learning
habits of the college students from Mumbai. He
may select the sample as under
1)First prepare a list of all colleges in Mumbai city
2)Then, select a sample of colleges on random
basis. Suppose there are 200 colleges in Mumbai,
then he may select 20 colleges by random
method.
81. DR G K KALKOTI81
3)From the 20 sampled colleges, prepare
a list of all students. From these lists
select the required number of say 1000
students on random basis]
In this example the researcher gets a
sample 1000 students from 20 colleges
only otherwise if researcher decides to
select 1000 students on random basis,
then he would have to select them out of
200 colleges which would have been
expensive and time consuming
82. Cluster Sampling
âĸ The target population is first divided into mutually exclusive and
collectively exhaustive subpopulations, or clusters.
âĸ Then a random sample of clusters is selected, based on a
probability sampling technique such as SRS.
âĸ For each selected cluster, either all the elements are included in
the sample (one-stage) or a sample of elements is drawn
probabilistically (two-stage).
âĸ Elements within a cluster should be as heterogeneous as
possible, but clusters themselves should be as homogeneous
as possible. Ideally, each cluster should be a small-scale
representation of the population.
âĸ In probability proportionate to size sampling, the clusters are
sampled with probability proportional to size. In the second
stage, the probability of selecting a sampling unit in a selected
cluster varies inversely with the size of the cluster.
83. Types of Cluster Sampling
Cluster Sampling
One-Stage
Sampling
Multistage
Sampling
Two-Stage
Sampling
Simple Cluster
Sampling
Probability
Proportionate
to Size Sampling
84. The population is divided into subgroups (clusters) like
families. A simple random sample is taken of the subgroups
and then all members of the cluster selected are surveyed.
Cluster Sampling
90. DR G K KALKOTI90
Multistage and Multi Phase
Sampling
ī As the name suggests, multistage sampling is
carried out in steps. This method is regularly
used in conducting national surveys on large
scale. It is an economical and time saving method
of selecting a sample out of widely spread
population.
ī In this method first the population will be divided
on state basis, then districts, then cities, then
locality, wards, individuals who are sampled at
different stages until a final sample unit.
91. DR G K KALKOTI91
Multiphase sampling is slightly different
from multi-stage sampling. With multi-
phase sampling, the sampling unit at
each phase is the same, but some of
them are interviewed in detail or asked
more questions than others ask. In other
words, all the members of the sample
provide basic information and some of
them provide more and detailed
information.
94. Non Probability Sampling
īļ Involves non random methods in selection of sample
īļAll have not equal chance of being selected
īļSelection depend upon situation
īļConsiderably less expensive
īļConvenient
īļSample chosen in many ways
95. DR G K KALKOTI95
Non-Probability Sampling
ī Non-probability sampling is also called as
judgment sampling.
ī In case of non-probability sampling, units in
the population do not have an equal chance
or opportunity of being selected in the
sample. The non-probability method believes
in selecting the sample by choice and not by
chance.
ī Non-probability sampling suffers defects like
personal bias and sampling error cannot be
estimated.
ī This is an unscientific and less accurate
method of sampling, hence it is only
occasionally used in research activities.
97. Features of the nonprobability
sampling
3/2/2015jaympatidar@yahoo.in97
98. Uses of Nonprobability Sampling
3/2/201598
ī This type of sampling can be used when
demonstrating that a particular trait exists in the
population.
ī It can also be used when researcher aims to do a
qualitative, pilot , or exploratory study.
ī It can be used when randomization is not possible
like when the population is almost limitless.
ī it can be used when the research does not aim to
generate results that will be used to create
generalizations.
ī It is also useful when the researcher has limited
budget, time, & workforce.
ī This technique can also be used in an initial study
(pilot study)
100. Purposive Sampling
3/2/2015100
ī It is more commonly known as âjudgmentalâ or
âauthoritative samplingâ.
ī In this type of sampling, subjects are chosen to be
part of the sample with a specific purpose in mind.
ī In purposive sampling, the researcher believes that
some subjects are fit for research compared to
other individual. This is the reason why they are
purposively chosen as subject.
ī In this sampling technique, samples are chosen by
choice not by chance, through a judgment made
the researcher based on his or her knowledge about
the population
101. CountâĻ
3/2/2015101
ī For example, a researcher wants to study the lived
experiences of postdisaster depression among
people living in earthquake affected areas of
Gujarat.
ī In this case, a purposive sampling technique is
used to select the subjects who were the victims of
the earthquake disaster & have suffered
postdisaster depression living in earthquake-
affected areas of Gujarat.
ī In this study, the researcher selected only those
people who fulfill the criteria as well as particular
subjects that are the typical & representative part
of population as per the knowledge of the
102. DR G K KALKOTI102
Purposive Sampling
īPurposive sampling means deliberate
selection of sample units confirm to some
predetermined criteria. This is also known
as judgment sampling
īIt involves selection of cases when we
judge as most appropriate ones for a given
study. It is based on the judgment of a
researcher. It does not aim at securing a
cross section of a population. The
selection of samples depends upon the
subjective judgment of researcher.
104. Convenience Sampling
3/2/2015104
ī It is probably the most common of all sampling
techniques because it is fast, inexpensive, easy, &
the subject are readily available.
ī It is a nonprobability sampling technique where
subjects are selected because of their convenient
accessibility & proximity to the researcher.
ī The subjects are selected just because they are
easiest to recruit for the study & the researcher did
not consider selecting subjects that are
representative of the entire population
ī It is also known as an accidental sampling.
ī Subjects are chosen simply because they are easy
106. CONVENIENCE SAMPLING
īļSometimes known as grab or opportunity
sampling or accidental or haphazard sampling.
īļA type of non probability sampling which
involves the sample being drawn from that part of
the population which is close to hand. That is,
readily available and convenient.
107. DR G K KALKOTI107
Convenience Sampling
ī In convenience sampling, the sample is selected
as per the convenience of the researcher.
ī For example, the producer may add a reply coupon
along with product to collect responses from
consumers. The duly returned coupons are
conveniently available to the researcher for the
survey purpose.
ī Manufacturers of consumer goods like Titan
watches and Philips provide a questionnaire along
with the product purchased and collect information
relating to name of retail store, income group etc.,
similarly sample selected from the telephone
directory, pay-roll register, register of members is
a type of convenience sampling.
108. Merits and Demerits
3/2/2015108
Merits
ī This technique is
considered
easiest, cheapest,
& least time
consuming.
ī This sampling
technique may
help in saving
time, money, &
resources.
Demerits
ī Sampling bias, & the
sample is not
representative of the entire
population.
ī It does not provide the
representative sample
from the population of the
study.
ī Findings generated from
these sampling cannot be
generalized on the
109. Consecutive Sampling
3/2/2015109
ī It is very similar to convenience sampling except
that it seeks to include all accessible subjects as
part of the sample.
ī This nonprobability sampling technique can be
considered as the best of all nonprobability
samples because it include all the subjects that
are available, which makes the sample a better
representation of the entire population.
110. 3/2/2015110
CountâĻ
ī In this sampling technique, the investigator pick
up all the available subjects who are meeting the
preset inclusion & exclusion criteria.
ī This technique is generally used in small-sized
populations.
ī For example, if a researcher wants to study the
activity pattern of postkidney-transplant patient,
he can selects all the postkideney transplant
patients who meet the designed inclusion &
exclusion criteria, & who are admitted in post-
transplant ward during a specific time period.
111. Merits and Demerits
3/2/2015111
Merits
ī Little effort for
sampling
ī It is not expensive,
not time
consuming, & not
workforce
intensive.
ī Ensures more
representativeness
of the selected
sample.
Demerits
ī Researcher has not set
plans about the sample
size & sampling schedule.
ī It always does not
guarantee the selection of
representative sample.
ī Results from this sampling
technique cannot be used
to create conclusions &
interpretations pertaining to
the entire population.
112. Quota Sampling
3/2/2015112
ī It is nonprobability sampling technique wherein
the researcher ensures equal or proportionate
representation of subjects, depending on which
trait is considered as the basis of the quota.
ī The bases of the quota are usually age, gender,
education, race, religion, & socio-economic
status.
ī For example, if the basis of the quota is college
level & the research needs equal representation,
with a sample size of 100, he must select 25 first-
year students, another 25 second-year students,
25 third-year, & 25 fourth-year students.
113. DR G K KALKOTI113
Quota Sampling
ī Quota sampling is the frequently used method of
sampling in marketing research. The basic
objective of quota sampling is to control biases
arising out of non-probability method by
stratification and the setting of quotas for each
stratum.
ī For instance, a sample of 40 students can be
selected from a group of 200 students
comprising of 120 boys and 80 girls. To make the
sample representative, the group of 40 should
include 24 boys and 16 girls (i.e. 120: 80 = 3: 2).
ī Quota sampling offer benefits of speed, economy
and simplicity. It is widely used in market
surveys and public opinion polls.
114. Merits and Demerits
3/2/2015114
Merits
ī Economically cheap,
as there is no need
to approach all the
candidates.
ī Suitable for studies
where the fieldwork
has to be carried out,
like studies related to
market & public
opinion polls.
Demerits
ī It not represent all
population
ī In the process of sampling
these subgroups, other
traits in the sample may be
overrepresented.
ī Not possible to estimate
errors.
ī Bias is possible, as
investigator/interviewer can
select persons known to
115. Snowball Sampling
3/2/2015115
ī It is a nonprobability sampling technique that is
used by researchers to identify potential
subjects in studies where subjects are hard to
locate such as commercial sex workers, drug
abusers, etc.
ī For example, a researcher wants to conduct a
study on the prevalence of HIV/AIDS among
commercial sex workers.
ī In this situation, snowball sampling is the best
choice for such studies to select a sample.
ī This type of sampling technique works like chain
referral. Therefore it is also known as chain
116. 3/2/2015116
CountâĻ
ī After observing the initial subject, the
researcher asks for assistance from the subject
to help in identify people with a similar trait of
interest
ī The process of snowball sampling is much like
asking subjects to nominate another person
with the same trait.
ī The researcher then observes the nominated
subjects & continues in the same way until
obtaining sufficient number of subjects.
117. SNOWBALL SAMPLING
īļ Selection of additional respondents is
based on referrals from the initial
respondents.
- friends of friends
īļ Used to sample from low incidence or rare
populations.
118. Merits and Demerits
3/2/2015118
Merits
ī The chain referral process
allows the researcher to
reach populations that are
difficult to sample when
using other sampling
methods.
ī The process is cheap,
simple, & cost-efficient.
ī Need little planning &
lesser workforce
Demerits
ī Researcher has little
control over the
sampling method.
ī Representativeness of
the sample is not
guaranteed.
ī Sampling bias is also a
fear of researchers
when using this
sampling technique.
119. PROBLEMS OF SAMPLING
3/2/2015119
ī Sampling errors
ī Lack of sample representativeness
ī Difficulty in estimation of sample size
ī Lack of knowledge about the sampling
process
ī Lack of resources
ī Lack of cooperation
ī Lack of existing appropriate sampling
frames for larger population
120. In A Nut Shell
ī Probability Sampling
- Simple Random â Selection at Random
- Systematic â Selecting every nth case
- Stratified â Sampling w/n groups of Populn
- Cluster â Surveying whole clusters of P/n
- Multistage â Sub samples from large smpl
121. ī Non- Probability Sampling
- Accidental â Sampling those most convnt
- Voluntary â Sample is self selected
- Purposive â Handpicking typical cases
- Quota â Sampling w/n groups of Ppln
- Snowball â building sample thru informnts
122. Sampling Error
Sampling error refers to differences between the
sample and the population that exist only because
of the observations that happened to be selected for
the sample
Increasing the sample size will reduce this type of
error
125. Sample Errors
īļError caused by the act of taking a sample
īļThey cause sample results to be different from the results of
census
īļDifferences between the sample and the population that exist
only because of the observations that happened to be selected for
the sample
īļStatistical Errors are sample error
īļWe have no control over
127. Non Response Error
A non-response error occurs when units
selected as part of the sampling procedure do
not respond in whole or in part
128. Response Errors
īļRespondent error (e.g., lying, forgetting, etc.)
īļInterviewer bias
īļRecording errors
īļPoorly designed questionnaires
īļMeasurement error
A response or data error is any systematic bias that occurs
during data collection, analysis or interpretation
129. Respondent error
īļ respondent gives an incorrect answer, e.g. due to prestige or competence
implications, or due to sensitivity or social undesirability of question
īļ respondent misunderstands the requirements
īļ lack of motivation to give an accurate answer
īļ âlazyâ respondent gives an âaverageâ answer
īļ question requires memory/recall
īļ proxy respondents are used, i.e. taking answers from someone other than
the respondent
130. Interviewer bias
īļ Different interviewers administer a survey in different ways
īļ Differences occur in reactions of respondents to different
interviewers, e.g. to interviewers of their own sex or own ethnic
group
īļ Inadequate training of interviewers
īļ Inadequate attention to the selection of interviewers
īļ There is too high a workload for the interviewer
131. Measurement Error
īļ The question is unclear, ambiguous or difficult to answer
īļ The list of possible answers suggested in the recording instrument
is incomplete
īļ Requested information assumes a framework unfamiliar to the
respondent
īļ The definitions used by the survey are different from those used by
the respondent (e.g. how many part-time employees do you have?
See next slide for an example)
132. Key Points on Errors
Non-sampling errors are inevitable in production of national
statistics. Important that:-
ī¯ At planning stage, all potential non-sampling errors are listed and steps
taken to minimise them are considered.
ī¯ If data are collected from other sources, question procedures adopted
for data collection, and data verification at each step of the data chain.
ī¯ Critically view the data collected and attempt to resolve queries
immediately they arise.
ī¯ Document sources of non-sampling errors so that results presented can
be interpreted meaningfully.