UNIT 3 Principles of Sampling
UNIT 3 Principles of Sampling
UNIT 3 Principles of Sampling
Principles of Sampling
POPULATION AND SAMPLES •
A.Y 2022-2023
BIOSTATISTICS
BSML-2108
TARGET POPULATION – known as the ideal population
we would like to describe
POPULATION • SAMPLING POPULATION – the population from which a
sample will actually be taken
➢ The set of all objects that possess some common set of
• STUDY UNITS or SAMPLING UNITS – individual elements
characteristics with respect to a research problem
in the population of interest
➢ The hypothetical (and usually) infinite number of
people to whom you wish to generalize. “The purpose of sampling in research to find
representative samples to avoid bias”
Two methods of obtaining information on the population:
WHY WE DO SAMPLING?
CENSUS SAMPLING
The process of obtaining ➢ Difficult to obtain data of the entire population given
Surveying only a sample
information about the
of the whole population to a certain period of time required by the study
population by including
make inference about ➢ Very expensive to study a whole population
every element of the
the population ➢ Some populations especially animals are constantly
population
moving, being born and dying
➢ Some items are destroyed during the process of
SAMPLE obtaining data
➢ Studying an entire population requires highly qualified
➢ Describes those individuals who are in the study expert and staff which may not always be available
➢ A small subset of the population for the study.
A substitute teacher wants to know how students in the HOW LARGE SHOULD THE SAMPLE BE?
class did on their last test. The teacher asks the 10
students sitting in the front row to state their latest test ➢ If repeated samples of the known sampling elements
score. He concludes from their report that the class did were taken, then the resulting sampling distribution
extremely well. would be a normal distribution.
CONFIDENCE INTERVALS
EXAMPLE:
SAMPLING PROCESS
MARGIN OF ERROR
Identifying target population
• Look at the research objectives
➢ How close the estimate from your sample should be to
• Consider the appropriate sampling unit
the actual population value • Specify clearly what is excluded
EXAMPLE: • Don’t over define
• Consider convenience
A newspaper recently reported one candidate for mayor ➢ SAMPLING FRAME – collection of the sampling units
was favored with 63% ± 3% of the vote
Determining sample frame
➢ The higher the level of confidence desired, the larger • Does not have to enumerate all population
the sample size members
➢ The more precise the required sample results, the • Subset problems
larger sample size • Superset problems
• Intersection problems
HOW TO COMPUTE FOR SAMPLE SIZE Selecting a Sample Procedure
• Non-probability vs Probability sampling
SLOVEN’S/SLOVIN’S FORMULA
NON-PROBABILITY SAMPLING TECHNIQUE
N = population
➢ Refer on the subjective judgment of the researcher
𝑁𝑒 2 = Sampling error/ when selecting units from the population to be
margin of error included in the sample
➢ The reliability of the resulting estimates cannot be
assessed
COCHRAN FORMULA
Judgmental/ Snowball Convenience Quota
Purposive sampling sampling sampling
➢ Used for a very large population
sampling
Where:
USE CASE:
STRATIFIED SAMPLING
➢ WHEN: Your given population is logically homogenous ➢ A method of selecting a sample in two or more steps
➢ HOW: In a systematic sample, after you decide the ➢ Population is divided into PSU’s – draw a sample –
sample size, arrange the elements of the population in draw SSU – and so on..
some order and select terms at regular intervals from ➢ Dividing the population into a hierarchy of sampling
the list units corresponding to the different sampling stage
➢ PROS: Simple. The assurance that the population will APPLICATION OF SAMPLING IN HEALTH RESEARCH
be evenly sampled
Sampling of non-human
➢ CONS: inherent periodicity of the list. This can be Sampling of communities
populations
avoided by randomizing the list of your population
entities.
USE CASE:
CLUSTER SAMPLING
EXAMPLE:
CLUSTER vs STRATIFIED
Cluster Sampling Stratified Random
Elements are Sampling
HETEROGENOUS Elements are
HOMOGENOUS
Clusters are
HOMOGENOUS Clusters/strata are
HETEROGENOUS