Sampling

What is sampling?

Sampling involves the selection of a number of study units from a defined study population.

The population is too large for us to consider collecting information from all its members.

Instead we select a sample of individuals hoping that the sample is representative of the

population.

When taking a sample, we will be confronted with the following questions:

a) What is the group of people from which we want to draw a sample?

b) How many people do we need in our sample?

c) How will these people be selected?

Definitions

Target population (reference population): Is that population about which an

investigator wishes to draw a conclusion.

Study population (population sampled): Population from which the sample actually

was drawn and about which a conclusion can be made. For Practical reasons the

study population is often more limited than the target population. In some

instances, the target population and the population sampled are identical.

Sampling unit: The unit of selection in the sampling process. For example, in a

sample of districts, the sampling unit is a district; in a sample of persons, a person,

etc.

Study unit: The unit on which the observations will be collected. For example,

persons in a study of disease prevalence, or households, in a study of family size.

N.B. The sampling unit is not necessarily the same as the study unit.

Sample design: The scheme for selecting the sampling units from the study population.

Sampling frame: The list of units from which the sample is to be selected.

40

Research methodology

The existence of an adequate and up-to-date sampling frame often

defines the study population.

Sampling methods

An important issue influencing the choice of the most appropriate sampling method is

whether a sampling frame is available, that is, a listing of all the units that compose the study

population.

a) Non-probability sampling methods

Examples:

1. Convenience sampling: is a method in which for convenience sake the study

units that happen to be available at the time of data collection are selected.
2. Quota sampling: is a method that insures that a certain number of sample units

from different categories with specific characteristics appear in the sample so

that all these characteristics are represented. In this method the investigator

interviews as many people in each category of study unit as he can find until he

has filled his quota.

3. Purposeful sampling strategies for qualitative studies: Qualitative research

methods are typically used when focusing on a limited number of informants,

whom we select strategically so that their in-depth information will give optimal

insight into an issue about which little is known. This is called purposeful

sampling.

The above sampling methods do not claim to be representative of the entire

population.

41

Research methodology

Random sampling strategies to collect quantitative data: If the aim of a study is to

measure variables distributed in a population (e.g., diseases) or to test hypotheses about

which factors are contributing significantly to a certain problem, we have to be sure that we

can generalise the findings obtained from a sample to the total study population. Then,

purposeful sampling methods are inadequate, and probability or random sampling methods

have to be used.

b) Probability sampling methods: They involve random selection procedures to ensure

that each unit of the sample is chosen on the basis of chance. All units of the study

population should have an equal or at least a known chance of being included in the

sample.

1. Simple Random Sampling (SRS): This is the most basic scheme of random

sampling. To select a simple

random sample you need to:

Make a numbered list of all the units in the population from which you

want to draw a sample. Each unit on the list should be numbered in

sequence from 1 to N (Where N is the Size of the population).

Decide on the size of the sample

Select the required number of sampling units, using a “lottery” method

or a table of random numbers.

2. Systematic Sampling: Individuals are chosen at regular intervals (for example,

every 5th, 10th, etc.) from the sampling frame. Ideally we randomly select a number

to tell us where to start selecting individuals from the list. For example, a

systematic sample is to be selected from 1000 students of a school. The sample

size is decided to be 100. The sampling fraction is: 100/1000 = 1/10. The number

of the first student to be included in the sample is chosen randomly by picking one

out of the first ten pieces of paper, numbered 1 to 10. If number 5 is picked, every

tenth student will be included in the sample, starting with student number 5, until

100 students are selected. Students with the following numbers will be included in

the sample: 5,15, 25, 35,45, . . . , 985, 995.

42

Research methodology

Systematic Sampling is usually less time consuming and easier to

perform than SRS.

It provides a good approximation to SRS.

Should not be used if there is any sort of cyclic pattern in the ordering

of the subjects on the list.

Unlike SRS, systematic sampling can be conducted without a sampling

frame (useful in some situations where a sampling frame is not readily

available).

4. Stratified sampling: If it is important that the sample includes representative

groups of study units with specific characteristics (for example, residents from

urban and rural areas), then the sampling frame must be divided into groups, or

strata, according to these characteristics. Random or systematic samples of a

predetermined size will then have to be obtained from each group (stratum).

This is called stratified sampling.

Some of the reasons for stratifying the population may be:

Different sampling schemes may be used in different strata, e.g. Urban

and rural

Conditions may suggest that prevalence rates will vary between strata: the

overall estimate for the whole population will be more precise if

stratification is used.

Administrative reasons may make it easier to carry out the survey through

an organization with a regional structure.

5. Cluster sampling: When a list of groupings of study units is available (e.g.

villages, etc.) or can be easily compiled, a number of these groupings can be

randomly selected. The selection of groups of study units (clusters) instead of

the selection of study units individually is called cluster sampling. Clusters are

often geographic units (e.g. districts, villages) or organizational units (e.g.

clinics).

43
Research methodology

6. Multi-Stage Sampling: This method is appropriate when the population is large

and widely scattered. The number of stages of sampling is the number of times

a sampling procedure is carried out.

The primary sampling unit (PSU) is the sampling unit (or unit of

selection in the sampling procedure) in the first sampling stage;

• The secondary sampling unit (SSU) is the sampling unit in the second sampling stage,

etc.

e.g.

After selection of a sample of clusters (e.g. household), further sampling of

individuals may be carried out within each household selected. This constitutes two

stage sampling, with the PSU being households and the SSU being individuals.

Advantages: less costly, we only need to draw up a list of individuals in the clusters

actually selected, and we can do that when we arrive there.

Disadvantage: less precise than SRS.

a) Sampling error (i.e., random error)

When we take a sample, our results will not exactly equal the correct results for the whole

population. That is, our results will be subject to errors. This error has two components:

sampling and non-sampling errors.

Random error, the opposite of reliability (i.e., Precision or repeatability), consists of random

deviations from the true value, which can occur in any direction.

Sampling error (random error) can be minimized by increasing the size of the sample.

Reliability (or precision): This refers to the repeatability of a measure, i.e., the degree of

closeness between repeated measurement of the same value. Reliability addresses the

question, if the same thing is measured several times, how close are the measurements to

each other?

44

Research methodology

The sources of variation resulting in poor reliability include:

a) Variation in the characteristic of the subject being measured. Example: blood pressure

b) The measuring instruments, e.g. questionnaires

c) The persons collecting the information (observer variation)

Inter-observer variation: differences between observers in measuring the same

observation

Intra-observer variation: differences in measuring the same observation by the same

observer on different occasions.

b) Non Sampling error (i.e., bias)

Bias, the opposite of validity, consists of systematic deviations from the true value, always in

the same direction.

It is possible to eliminate or reduce the non-sampling error (bias) by careful design of the

sampling procedure.

Validity: This refers to the degree of closeness between a measurement and the true value

of what is being measured. Validity addresses the question, how close is the measured

value to the true value?

To be accurate, a measuring device must be both valid and reliable. However, if one cannot

have both, validity is more important in situations when we are interested in the absolute

value of what is being measured. Reliability on the other hand is more important when it is

not essential to know the absolute value, but rather we are interested in finding out if there is

a trend, or to rank values.

Examples of types of bias in sampling include:

Bias resulting from incompleteness of the sampling frame: accessibility bias, seasonability

bias, self-reporting bias, volunteer bias, non-response bias etc.

45

Research methodology

Non-response bias refers to failure to obtain information on some of the subjects included in

the sample to be studied. It results in significant bias when the following two situations are

both fulfilled.

1. When non-respondents constitute a significant proportion of the sample.

2. When non-respondents differ significantly from respondents.

The issue of non-response should be considered during the planning stage of the

study:

a) Non-response should be kept to a minimum. E.g. below 15%

Methods that may help in maintaining non-response at a low level could be:

• Training data collectors to initiate contact with study subjects in a respectful way and

convince them about the importance of the given study (this minimizes the refusal type

of non-response)

• Offering incentives to encourage participation (this should be done by taking account

of the potential problems that may arise in conducting future research)

• By making repeated attempts (at least 3 times) to contact study subjects who were

absent at the time of the initial visit.

b) The number of non-responses should be documented according to type, so as to

facilitate an assessment of the extent of bias introduced by non-response.

c) As much information as possible should be collected on non-respondents, so as to see

in what ways they may differ from respondents.

• Selection bias cannot be corrected by increasing the size of the sample, why? How

do you remove this type of bias?