Sampling

9.
Sampling
Basic Sampling Concept

Definition
• This is the method or process of selecting
certain members, objects, events to
represent the whole group
• But this is not possible due to:
• Time factor
• Cost
• Manageability
Basic Concept in Sampling
• 1. Population: this is the entire group of
persons or objects and events of interest to the
researcher ,
• stipulate inclusion criteria ie eligible criteria, or
distinguishing descriptors
• The population the researcher defines is
referred to as accessible population or study
population
Concepts cont’
• 2. Elements: this is the unit that make up
the total population and from which data is
obtained. This could be people, objects or
events. ie a unit of analysis
• 3. Parameter: a characteristic of the
elements of a population such as the
mean age. This describes a particular
characteristics of the whole population
• 4. Sample: subset of a population
Concepts cont’
• 5. Sampling frame:
• 6 Representative sample: This is a sample
that resembles the population in as many
ways as possible enabling the researcher
to generalise the results: issues to note
are demographic characteristics eg age,
ethnicity etc
Concepts cont’
• 7. Sampling error: this refers to the difference between
the population parameters and sampling statistics.
Although this is unavoidable but a large sampling error
means the sample is not representative of the
population this can be due to the small size of the
sample. Sampling error are due to:
• Chance factor: this is the inclusion of an elements at
the expense of the other and this can be statistically
calculated
• Bias selection: due to over or under representation
• Non-responsive error: for some reason a
respondent does not respond to a measuring
instrument and hence is excluded
Concept cont’
• 8. Sampling bias: this occurs when there
is a difference between sample data and
population data.
• Sources of sampling bias
– Language barrier
– Personal view
– Place of data collection
Sampling approaches
• Probability: this is an approach that gives
each element an equal chance of being
chosen in the sample.
• probability sampling is divided into the
following:
• Random sampling
• Systematic sampling
• Stratified random sampling
• Cluster sampling
Probability Sampling
• Random sampling: This is the most basic approach to
probability sampling that gives all elements in a
population equal chance based on the definition of the
population and sample frame
• Once sample size has been determined allocate
consecutive identification numbers to the elements then
decide on the random selection technique and the
common ones being:
• Fishbowl or lottery: all elements identified and placed in
fish bowl and drawn randomly
• Random number table: pick starting point in the table
and then move horizontally, vertically and then
diagonally each time picking numbers that are in the
population. If number occurs that is not in the
represented population exclude and move on
Probability Sampling cont’
• Systematic sampling: sometimes referred to as
interval sampling where every fth of the
population is selected eg this is simpler and
quicker but depends on the availability of a
complete population list same as random
• The process involves
– Obtain list of the population (N)
– Determine sample size (n)
– Determine the sample interval by dividing the size of
the population by the size of the sample
Probability Sampling cont’
• Stratified sampling: divided into
• Stratified Random Sampling: This is used when
the population is divided into strata ie age and
then selection from each is done using either
random or systematic
• Proportional stratified sample: the sample must
reflect the distribution of the elements to avoid
over sampling one group or event or object. This
requires extensive knowledge of the population
• Disproportional stratified sample: ideal when
finding are not to be generalised across the
group.
Probability Sampling Cont’
• Cluster sampling: this is when the population is
grouped into heterogeneous cluster as
compared to homogenous ones. Ideal when the
whole list of the sample frame is not available or
a large group is involved
• Cluster sampling takes place in stages eg start
with general population then cities, then
progress to residential area and then individual
participants
• Although it tends to contain more error than the
other two it is widely used when the population is
large and widely scattered
Non probability sampling
• This is the method where the researcher
does not know whether all the elements of
the population are included or can be
included giving a zero chance to some
elements
• This is less desirable since it is next to
impossible to generalise the findings
• Used when probability sampling is difficult
to implement or too expensive
Non probability sampling cont’
• Types
• 1. Convenience sampling: This is based on
availability of elements chosen based on researcher’s
convenience: problem – overrepresentation or under
• 2. Quota sampling: this is associated with stratified
sampling but elements chosen based on convenience
• 3. Purposive or judgemental: used when researcher
has sufficient knowledge of the population e.g. days of
high and low activity
• 4.Snowball sampling: used when it is difficult to
locate element hence once few are found then these
are used to locate and identify others

Sampling

Uploaded by

Copyright:

Available Formats

Sampling

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Sampling

Uploaded by

Copyright:

Available Formats

9.

Basic Sampling Concept

You might also like