SAMPLING TECHNIQUES

Biostatistics Course 2021-2022 / Block 5

Ali Lateef Jasim

MBChB.
Learning objectives

❑ Define statistical inference, population of units, population of

observations, sample of units, sample of observations, sampling
error and sampling frame.
❑ List the types of samples and Discuss the sampling technique of
each.
❑ Identify the differences between probability and non-probability
samples.
❑ Appraise the advantages and disadvantages of sampling.
 Definitions
❖ Statistical inference: is a conclusion concerning a population
of observations (or units) made on the basis of the results of
studying a sample of these observations (or units).
❖ Population of units: a group of entities that have
quantifiable characteristics in common e.g.: population of
human being, animals, bacteria, machines …etc.
❖ Population of observations: a group of numerical values of
quantifiable characteristics of each member of population
of units e.g.: list of BP values of certain population.
❖ Sample of units: a finite number of units from a population
of units.
❖ Sample of observations: a finite number of observations
from a population of observation.
 Definitions
❖ A Population: a set which includes all measurements of
interest to the researcher (The collection of all responses,
measurements, or counts that are of interest).
❖ A Sample: a subset of the population.
❖ Sampling Error: it is the difference between a sample
measure and its corresponding population measure.
✓ Sampling error is not a mistake, but it is a calculated
error that should be quantify.
❖ Sampling frame: It is a numerical list of all the units
composing the study population i.e. every member of the
population has a unique identification number.
 Statistical inference

❖ In inferential statistics, the data are taken from the sample

and allows you to generalize the population.
❖ In general, inference means “guess”, which means making
inference about something.
❖ The purpose of statistical inference to estimate the
uncertainty or sample to sample variation.
❖ It allows us to provide a probable range of values for the
true values of something in the population.
 Inferential Statistics
Sample

Sample

Population Sample

Sample

Draw inferences about

the larger group
 Types of sampling methods

❖ Probability sampling:
✓ The results of studying the sample are generalizable to
the underlying population from which this sample had
been drawn.
❖ Non-probability sampling:
✓ The results of studying the sample can not be
generalized to the underlying population from which
this sample had been drawn.
 Probability (Random) Sample
❖ The sample is drawn from the population in such a way
that every member of the population has the same
probability (chance) to be included in the sample.
❖ Each subject has a known probability of being selected
❖ Helps in: ✓ Generalizing the result.
✓ Testing hypotheses.
❖ Types 1. Simple random sample.
2. Systematic Sample.
3. Stratified sample.
4. Cluster sample.
 Types of probability samples

1. Simple random sample: It requires:

❖ Sample frame: a numerical list of all observations (or units)
composing the population.
✓ Examples: Map, Census database, Employee database,
Telephone directory.
❖ Sample fraction: sample size to the total population.
 1. Simple random sample
Selection of Simple Random Sample from the population can
be done by:
✓ Lottery method.
✓ Computer generated random sampling (random number
generator).
✓ Many available online sites: e.g. www.random.org/integers.
✓ Use Excel: E.g. “=randbetween(1,200)” generates a
random number between 1 and 200.
✓ Random number table (random digit).
Simple random sample
 How to use the random number table?

1. Let’s assume that we have a population of 200 students

and each student has been assigned a number from 1 to
200. Suppose we wish to sample 5 students.
2. Since we have a population of 200 and 200 is a three digit
number, we need to use the first three digits of the numbers
listed on the chart.
3. We close our eyes and randomly point to a spot on the
chart. For this example, we will assume that we selected
2134 in the last column.
 How to use the random number table?

4. As we take only the first 3 digits, then it will be 213 (but 213
is more than 200, so we move to the next number in the
column).
5. Next number is 1095 (first 3 digits are 109), so the first
student to be selected is No. 109.
6. And we continue so on until we get the numbers of the 5
students we want to include in the study.
 2. Systematic Sample

❖ Choosing units or observations from the sample frame at

regular interval (every nth).
❖ To find the “system” we divide the population size by the
required sample size.
✓ e.g. if the population is composed of 300 units & we
want to select a sample of 100, we can use the interval:
✓ 300/100 = 3 , So we will choose every 3rd
❖ The starting point can be chosen at random like 2, so the
selected units will be 2, 5, 8, 11, …and so on.
2. Systematic Sample
2. Systematic Sample
 Limitations

❖ Simple random sample & systematic sample can't ensure

that the structure of the sample will be similar to the
structure of the underlying population regarding certain
characteristics, as age , sex, …
❖ Difficulty accessing lists of the Full Population.
 3. Stratified sample

❖ The sample frame is divided into strata (or groups)

according to certain characteristic (s), and then a simple
random or systematic sampling will be applied on each
stratum.
❖ The number of units included in the sample from each
stratum can be achieved by:
1. Equal allocation from each stratum.
2. Proportional allocation.
 3. Stratified sample

❖ Proportional allocation: The number from each stratum is

proportional to the size of the stratum, this is possible only if
we know the proportion of population in each stratum to
the whole study population.
n/N= n1/N1= n2/N2 … nk/Nk
Example: If one stratum consists of 1000 people, and another of
2000 people, In this method, the researcher draws the same
proportion of people from each stratum, such as 5% of the
stratum.
Proportional allocation
 Limitations

❖ This method's disadvantage is that several conditions must

be met for it to be used properly. Researchers must identify
every member of a population being studied and classify
each of them into one, and only one, subpopulation. As a
result, stratified random sampling is disadvantageous when
researchers can't confidently classify every member of the
population into a subgroup.
 4. Cluster sample

❖ The selection here will be of groups of units rather than

individual units.
❖ A sample frame of groups of study units (cluster) should be
available, then a random sample of these clusters will be
chosen.
❖ Cluster could be schools, clinics, hospitals, villages, factories…
etc.
❖ Example: To study certain characteristic in medical students,
here the clusters are the different colleges from which we
can select entire colleges (clusters) as a sample.
4. Cluster sample
4. Cluster sample
4. Cluster sample
 4. Cluster sample

Advantages:
1. Requires fewer resources.
2. More feasible.

Disadvantages:
1. Biased samples.
2. High sampling error.
 Multi-stage sample

❖ This procedure is carried out in phases (stages).

❖ It can involve more than one of the above sampling
method.
❖ It is used for very large population & when the sample
frame is not available for the whole population.
Types of non-probability samples

1. Convenience sample.
2. Purposive sampling (judgmental).
3. Quota sample.
4. Snowball sampling.
 Types of non-probability samples
1. Convenience sample:
✓ The study units is selected from elements of a population
that are easily accessible that happen to be present at the
time of data collection. This is not representative to the
population we want to study.
✓ Example: A university student working on a project and
wants to understand the average hours of sleep per day of
students will most possibly call his/her classmates and friends
and ask how many hours do they sleep per day as an
average. Or may go to a public place nearby and conduct
an easy survey.
 Types of non-probability samples

2. Purposive sampling (judgmental): You chose who you

think should be in the study.
3. Quota sample:
✓ The composition of the sample regarding certain
characteristic and the number of units having these
characteristics are decided from the beginning, and the only
requirement is to find the people fitting these quotas.
4. Snowball sampling: (friend of friend …etc.).
 Example of Quota sample
A researcher wants to survey individuals about what
smartphone brand they prefer to use.
He/she considers a sample size of 500 respondents. Also, he/she
is only interested in surveying ten states in the US. Here’s how
the researcher can divide the population by quotas:
✓ Gender: 250 males and 250 females.
✓ Age: 100 respondents each between the ages of 16-20, 21-
30, 31-40, 41-50, and 51+.
✓ Employment status: 350 employed and 150 unemployed
people.
✓ Location: 50 responses per state
 Snowball sampling

❖ Used when the desired sample characteristic is rare or when

locating respondents is extremely difficult or costly.
❖ The technique relies on referrals from initial subjects to
generate additional subjects.
❖ It lowers search costs; but the sample will not represent the
population.
❖ This sampling method involves a primary data source
nominating other potential data sources that will be able to
participate in the research studies.
Snowball sampling
 Properties of Snowball Sampling
Is useful when there is:
✓ No official list of names of the members.
✓ Difficulty to locate people.
✓ People who are not willing to be identified.

So, its advantages are:

1. It’s quicker to find samples.
2. Cost effective.
3. Sample hesitant subjects.
 Non- probability sampling

❖ The results of non-probability sample are valid (to the

studied sample) but are not generalizable to the underlying
population.
❖ So non-probability sample are inappropriate if the aim is
to generalize finding obtained from the sample to the total
study population.
❖ Researchers use this method in studies where it is impossible
to draw random probability sampling due to time or cost
considerations.
 Advantages of collecting
information using samples

❖ Sampling may be the only feasible method of collecting the

information.
❖ Sampling reduce demand on resources (finance, personnel,
materials).
❖ Results are obtained more quickly.
❖ Sampling may lead to better accuracy of collected data.
❖ Precise allowance can be made for sampling errors.
 Disadvantages of collecting
information using samples

❖ There is always sampling error.

❖ Sampling may create a feeling of discrimination within the
population.
❖ Sampling may be inadvisable where every unit in the
population is legally required to have a record.
❖ For rare events, small samples may not yield sufficient cases
for study.
Thank You
Ali Lateef Jasim
MBChB.