Sampling Procedures

By
Dr. Mina Nakawuka
Preamble
• It is virtually impossible to study every individual in the target
population.

• e.g all clients receiving ART in Uganda, is simply too large for the
researcher to plan a quality research study.

• Collecting millions of questionnaires from every patient would

challenging
• Do researchers then have to restrict target population to small groups eg from
one ART clinic- so that the researcher can access the entire population? NO!

• "Paradoxically, the attempt to observe all cases [in a population] may actually
describe a population less accurately than a carefully selected sample... "
(Singleton & Straits, 2010, p. 151).

• Thus researchers have developed sampling procedures that should identify a

sample that is representative of the population

• Hence, a sample closely resembles the target population on all relevant

characteristics
Study objectives
By the end of the lecture, students should be able to:

• Define and justify the reasons for sampling

• Select and apply appropriate probability methods of sampling in a
given research
• Select and apply appropriate non-probability methods of sampling in
a given research
• Describe the different biases that are likely to occur in sampling and
how they can be minimized
What is sampling?
• Is a process or technique of selecting a representative a sub-group from a
population under the study.

Or
• It is the process of selecting a number of individuals for a study in such a way
that the individuals selected represent the large group from which they were
selected (Ogula, 2005).

•  How representative is the sample of the target population? 

• Population validity, the degree to which the results of a study can be

generalized from the sample to the target population.
Theory of Sampling

• The theory of sampling is as follows:

• Researchers want to gather information about a whole group of

people (the population).

• Researchers can only observe a part of the population (the sample).

• The findings from the sample are generalized, or extended, back to

the population.
Theory of sampling
Key definitions
• Population: refers to a large set or collection of items with
something in common

• Sample: group of people who take part in the investigation.

The people who take part are referred to as “participants”.

• Sampling unit: one of the units into which an aggregate is

divided for the purpose of sampling, each unit being
regarded as individual and indivisible when the selection is
made.
Definitions cont’d
• Sampling frame: list of items in your population

• Target population: group that we are interested in studying. . It the

population from which the sample will be drawn

• Sampled population or study population: this is the accessible

population
• Generalisability: extent to which we can apply the findings of our
research to the target population we are interested in.
Census
• When one wants to study the entire population, then he is dealing
with a census

• Under the census or complete enumeration survey method, data

are collected for each and every unit [person, household, field,
shop, factory etc.] of the population or universe
Census cont’d
Advantages of census
• Data obtained from each and every unit of the population
• The results obtained are likely to be more representative, accurate and reliable

Disadvantages of a census
 Costly
 Time consuming
 Sometimes not necessary

• That’s why normally samples are used

Importance of sampling in research
• 1. A sample can be studied more quickly than a population

• 2. It is less expensive

• 3. A census (study of the entire population) is impossible in most

situations

• 4. Usually, sample results are more accurate

Importance cont’d
• 5 If properly selected, probability methods can be used to estimate
the error

• 6. A sample of the population with more representative

characteristics is more appropriate than the entire population in
studying certain aspects of a disease or health related conditions e.g.
a study among a high-risk sub-population
• Disadvantage
• . In sample survey, more time and resources are spent in training
personnel for a smaller scope of work
Types of sampling procedures
• There are two major sampling procedures in research.

1) Probability procedures

2) Non probability sampling.

 Classification of Sampling Methods

Sampling
Methods

Probability Non-
Samples probability
Multistage

Systematic Convenience Snowball

Stratified

Simple
Cluster Judgment Quota
Random
1) Probability Sampling Procedures

• Involve random selection procedures to ensure that each unit of the

sample is chosen on the basis of chance.

• Every unit has an equal known chance (greater than zero) of being
selected in the sample. The best way to ensure that a sample will lead
to reliable and validly representative inferences is to use probability
samples .

• Every element is selected independently of every other elements

• It can be done by use of lottery method or random method

Types of probability Sampling Procedures
• Four basic types which include;

• i)simple random sampling (SRS),

• ii)systematic sampling,
• iii)stratified sampling
• iv)cluster sampling
• V) Multi stage sampling
i) Simple Random Sampling Procedure

• Provides the base for the other more complex sampling

methodologies

How to conduct SRS

• The researcher must first prepare the sampling frame of interest.

• Each element is accorded a serial number and the entire series is

called a ‘sampling frame’
SRS Cont’d
• The sample is then drawn so that each person or item has an equal chance of
being drawn during each selection round

• Random number tables generated by computerized sampling programs or random

number generator are used to select members of the population to be sampled.

• This gives each person/ member remaining in the population same probability of
being selected for the sample hence the randomness.

• Sampling can be ‘with replacement’ or ‘without replacement’.

• In the field, we may not use random numbers, but other field techniques that give
a semblance of randomness can be employed like lottery method e.g selecting
Village in a parish
SRS cont’d
• Advantages:
•  its ease of use 
• its accurate representation of the larger population

• Disadvantages:
• Tedious for large samples: time needed to gather the full list of a
specific population to make a sampling frame

• Bias could occur from larger population when sample is not

inclusive enough.
ii) Systematic Sampling Procedure
• Often used in place of simple random sampling.

• Researcher selects every nth member after randomly selecting the

first member laying between 1-nth element as the starting point.

• n is derived by dividing the number of items in the frame by the

desired sample size; e.g.

• E.g if the sample size is 20 respondents from a population of a 100,

every 5th member of the population will systematically be selected.
Cont’d
• To ensure randomness, the beginning number in the sampling frame
is selected using a table of random numbers or other probability
technique

• Advantages:
• Easier to draw and less time consuming ,
• It may spread the members selected for measurement more evenly
across the entire population than simple random sampling
iii) Stratified Sampling Procedure

• Most effective method to obtain representative sample

• It  involves categorizing the homogeneous members of the

population into mutually exclusive and collectively exhaustive
groups (strata) e.g sex male /female,

• An independent simple random sample is then drawn from each

group.
cont
• Advantages: Allows a relatively larger sample from a smaller population.

• Allows for subgroup analysis

• It also allows equitable representation of important strata hence able to give the
most  representative sample of a population

• Disadvantages: There is a potential for unequal sampling fractions; it is

important to ‘adjust’ for it when generalizing findings.

• it is unusable when researchers cannot confidently classify every member of the

population into a subgroup e.g race, ethnicity or religion into play
iv) Cluster Sampling Procedure

• In cluster sampling, a cluster (a group of population elements),

constitutes the sampling unit mainly geographically driven e.g
districts, sub-counties, villages and households

• Conducted due to cost efficiency  (economy and feasibility). The

sampling frame is also often readily available at cluster level and takes
short time for listing and implementation.

• Elements of a cluster however are not homogenous

• Advantages: Allows a relatively larger sample from a smaller population.
• It also allows equitable representation of important clusters

• Disadvantages: There is a potential for unequal sampling fractions; it is important to

‘adjust’ for it when generalizing findings.

• it may not reflect the diversity of the community; other elements in the same cluster may
share similar characteristics; 

• provides less information per observation than an SRS of the same size (redundant
information: similar information from the others in the cluster)

• standard errors of the estimates are high, compared to other sampling designs with the
same sample size.
v) multi-stage sampling
• Sampling is carried out in stages using smaller and smaller sampling units at each stage

• It’s a combination of methods.

• The sampling units are first the bigger units, either strata or clusters.
• Then we sample from a population of sub-clusters or sub-strata within the bigger units.
• It may sample individuals from the sub-clusters. The different levels of sampling are
referred to as stages.
• Multistage sampling usually involves a mix of different sampling methods at different
levels ie.
• Multi-stage cluster sampling
• b. Multi-stage stratified sampling
Cont’d
• Advantages:
• It allows a very large and diverse population to be studied.
• A sampling frame is not required for the entire population – we only
need a frame for the clusters or strata at the different stages.
• The sample is easier to select than a simple random sample of similar
size.

• Disadvantage:
• There is a larger probability of non-representativeness. We therefore
have to adjust for the ‘design effect’ and this increases the sample
size
2. Non Probability Sampling Procedures

• Probability that a person is selected is not known, they do not fulfill the requirements of
randomness needed to estimate sampling errors

• Used in situations where population may not be well defined.

• Often applied in qualitative studies but can also in quantitative methods when , there may not
be great interest in drawing inferences from the sample to the population
• Sometimes, we sample based on a purpose and sometimes not.

• They however often reflect biases (selection biases);

• The most common reason for using non probability sampling procedure is that it is less
expensive than probability sampling procedure and can often be implemented more quickly
(Michael, 2011).  

• It includes purposive, convenience and quota sampling procedures.

Cont’d
Why non-probability methods
• It is less expensive than probability sampling procedure
• Can often be implemented more quickly  

Types of non probability methods

• i)purposive,
• Ii)convenience and
• Iii) quota sampling procedures.
• Iv) snowball
Purposive/Judgmental Sampling Procedure

• Researcher chooses the sample based on who he/she thinks would be

appropriate for the study.

• The main objective of purposive sampling is to arrive as at a  sample

that can adequately answer the research objectives.

• The selection of a purposive sample often based on expert

knowledge is often accomplished by applying expert knowledge of the
target population
Cont’d
• Advantages: It may increase internal validity of findings and allows for exploratory depth.
Information may be useful for formation of hypotheses

• Disadvantage of purposive sampling

• It is open to bias (surveillance, selection and information biases) and findings may not be
generalizable

• Method is most appropriate for the selection of small samples often from a limited
geographic area or from a restricted population definition.

• The knowledge and experience of the researcher making the selections is a key aspect of
the ‘‘success’’ of the resulting sample (Michael, 2011).

• Key informants are also selected using this procedure.

ii)Convenience  Sampling Procedure

• Sometimes called opportunity, accidental or haphazard sampling.

• involves the sample being drawn from that part of the population which is
close to hand, that is, a population which is readily available and convenient.

• Often applied for pilot testing.

• Primary selection criterion relates to the ease of obtaining a sample ie cost

of locating elements of the population, the geographic distribution of the
sample, and obtaining the interview data from the selected element
• Advantage: Simplicity and resources

• Disadvantages: May not be representative of the population, hence

affecting generalisability
Quota sampling
• We allocate a quota to the different strata or clusters.

• The method ensures that a certain number of sample units from

different categories with specific characteristics appear in the sample,
so that all the categories are equitably represented.

 
Steps of quota sampling
• A population is first segmented into mutually exclusive sub-groups,

• Then judgment is used to select the subjects or units from each

segment based on a specified proportion without randomness.

• For example, interviewers might be tempted to interview those

people in the street who look most helpful, or may choose to
use accidental sampling to question those closest to them, for time-
keeping sake.
Advantage:
• Useful when probability samples are unlikely to provide the desired
balance of study units

Disadvantage:
• Non-representation and lack of generalisability
 iv)Snowball Sampling
• A snowball sample is achieved by asking a participant to suggest
someone willing or appropriate for the study.

• Snowball samples are particularly useful in hard-to-track populations

• E.g sex workers, LGBT
Note:
• Although these methods have less external validity, if done
correctly, and tagged to answer specific problems, they may have
a high internal validity.

• They are more robust when the main issue is exploratory depth
of the experiences of a few individuals known potentially to
harbour a ‘wealth of experience’
Bias in Sampling

Bias in sampling is a systematic error in sampling procedures that may lead to a

distortion in results of the study.

It includes:
(1) Non-response: Some sampled respondents fail to respond e.g In a survey (STIs),
many men refused to answer certain questions

(2) Self-selection bias (studying volunteers only or mainly). some members of the
intended population are less likely to be included than others
These may behave differently because of their inherent motivation

(3) Sampling registered patients only: They are likely to differ systematically from
the population e.g. some referral hospitals and units
Bias cont’d
(4) Missing Cases of short duration: In studies involving short duration
illness, some cases may be missed

(5) Seasonal Bias: Seasonal variations in the variable under study with
regard to incidence and prevalence e.g prevalence of malnutrition

(6) ‘Tarmac bias’: Selection of study areas and participants due to their
easy accessibility along main roads
Addressing Biases in Sampling

1) Data collection tools need to be simple

2) Pretesting of data collection tools

3) Follow up on non-response due to absence

4) Non response due to refusal to co-operate – an extra study for non respondents may
be considered; full information may facilitate consent. We can replace the non-
responders
5) Including additional people in the sample so that so absent are replaced

6) Anticipating the problem and solving it during the process or avoiding it

References

• http://www.kenpro.org/sampling-procedures/

• WHO http://apps.who.int/medicinedocs/en/d/Js6169e/7.5.html
Thank you for listening