Sampling Design

Designed with by

www.PresentationGO.com

The free PowerPoint template library

Objectives

The image part with relationship ID rId3 was not The image part with relationship ID rId5 was not The image part with relationship ID rId7 was not The image part with relationship ID rId9 was not
found in the file. found in the file. found in the file. found in the file.

1 2 3 4

Define Distinguish Compare Evaluate the

population, between sampling for appropriatene
sample and nonprobability quantitative ss of the
sampling and probability and sampling
samples and qualitative method and
compare their research sample size
advantages and methods used in
disadvantages studies
 Sampling

01 02 03 04 05

Sampling Population Element Sample Subject

Selecting sufficient An entire
number of elements Single Subgroup of Single member
group of the
from population so that
people, events
member of of sample
features of the sample
or things of population population
(e.g., mean) can be
generalized to the interest
population

www.presentationgo.com
 Population (target)

Accessible

Sample
Element

www.presentationgo.com
 Characteristics of a Good Sample
Design Marginal
Small sampling
systematic bias
error
Sampling error is reduced Avoid overrepresentation or
by selecting a large underrepresentation of some
sample and by using segment of the population;
efficient sample design decreases generalizability

01 and estimation strategies.

03 05
Representative 02 04
A small quantity of
something such as Sample should be
customers, data, people, Economically
viable generalized
products, or materials,
whose characteristics The sampling should be The sampling design should be
represent (as accurately as done in such a way that it is created keeping in mind that
possible) the entire batch, within the research budget samples that it covers the whole
lot, population, or and not too expensive to be universe of the study and is not
universe. replicated. limited to a part.

www.presentationgo.com
 Sampling Criteria
Exclusion
Inclusion criteria: criteria:
desirable undesirable
characteristics characteristics;
will not be
• Able to speak/read included
English

>>>>>>>>>
• have used an eBook • have not used an
in your subject area eBook in your
subject area
• 18 to 45 years old
• Age <18 years
>>>>>>>>>
• Unable to read or
speak English

www.presentationgo.com
 Variations in Sampling

1 Random variation 2 Systematic variation

(bias)
• Expected differences • Result of sampling
in values of a variable error
within a sample • Occurs when
• Variation around a sample is not
mean representative of
• Humans are unique population
and different!

www.presentationgo.com
Sampling Error
• Measured difference between the population mean
and the sample mean
• Representativeness of sample (the closer it is to the
population) minimizes sampling error

ERROR

Population Sample

Pop. mean Sample mean

www.presentationgo.com
Types of Systematic Variation
Acceptance rate:
Refusal rate: percentage of
percentage of subjects that
subjects that consented to be
declined to in the study
participate in the • 80 subjects
study approached
• 80 subjects and 76
approached accepted
and 4 refused • 76 80 = 0.95
• 4 80 = 0.05 (95%
(5% refusal acceptance
rate) rate)

www.presentationgo.com
Sampling Design Process
Define Population

Determine Sampling Frame

Determine Sampling Method

Probability Sampling Non-Probability Sampling

• Simple Random • Convenient
• Systematic • Judgmental
• Cluster • Snowball
• Stratified • Quota

Determine Appropriate Sample

Size

Execute Sampling Design

www.presentationgo.com
 Probability or Nonprobability

A sampling
02
technique in Nonprobability
which every
member of the 01
population will • Units of the sample are chosen on
have a known, Probability the basis of personal judgment or
nonzero convenience
• There are NO statistical
probability of techniques for measuring random
being selected sampling error in a non-
probability sample.
• Therefore, generalizability is never
statistically appropriate.

www.presentationgo.com
 Random Sampling Systematic Sampling
• Each member of the Random sampling is Select every Kth
population has an the purest form of individual from a list,
equal and known probability using a randomly selected
chance of being
selected.
sampling. starting point
Researcher must
• When there are very
know number of

01 02
large populations, it
is often ‘difficult’ to elements in the
identify every population and
member of the
the sample size
population, so the
pool of available desired
subjects becomes You can use software, such After the required sample
biased. as minitab to generate size has been calculated,
random numbers or to draw every Nth record is selected
directly from the columns from a list of population
members.

www.presentationgo.com
 Stratified Sampling
A stratum is a subset of the
population that share at
least one common
Stratified sampling is characteristic; such as males
commonly used and females.
probability method that
is superior to random The image part with relationship ID rId3 was not
found in the file.

• Identify relevant stratums and

sampling because it their actual representation in the
population.
reduces sampling error.
• Random sampling is then used to
select a sufficient number of
subjects from each stratum.
• Stratified sampling is often used
when one or more of the
stratums in the population have
a low incidence relative to the
other stratums.
www.presentationgo.com
www.presentationgo.com
 Cluster Sampling
The image part with relationship ID
rId7 was not found in the file.

The image part with relationship ID

rId3 was not found in the file.

Cluster Sample: a
probability sample in
which each sampling
Examples of
unit is a collection of clusters:
elements. • City blocks – political or
The image part with relationship ID
geographical
• Housing units – college
rId5 was not found in the file.

Effective under students

the following • Hospitals – illnesses
conditions: • Automobile – set of four
• A good sampling frame is not tires
available or costly, while a frame
listing clusters is easily obtained
• The cost of obtaining observations
increases as the distance separating
the elements increases
www.presentationgo.com
www.presentationgo.com
Purposive

Convenience
sampling is used in The image part with relationship
ID rId7 was not found in the file.
Selection
exploratory research based on
The image part with The image part with
relationship ID rId13 relationship ID rId9
was not found in the was not found in the
file. file.

where the
researcher is specific
interested in getting characteristics
an inexpensive
approximation.

The image part with relationship

ID rId3 was not found in the file.

The image part with relationship

ID rId5 was not found in the file.

Often used during preliminary

research efforts to get an estimate
without incurring the cost or time
The image part with
relationship ID rId11
was not found in the
file.

required to select a random sample

www.presentationgo.com
www.presentationgo.com
 Judgment Sampling

Judgmental sampling
A procedure in which a When using this method, is a non-probability
researcher exerts some the researcher must be sampling technique
effort in selecting a confident that the chosen where the researcher
sample that he or she sample is truly selects units to be
believes is most representative of the sampled based on their
appropriate for a study entire population. knowledge and
professional judgment

www.presentationgo.com
www.presentationgo.com
Quota Sampling

The image part with The image part with The image part with
relationship ID rId5 was not relationship ID rId7 was not relationship ID rId3 was not
found in the file. found in the file. found in the file.

A sampling method of
gathering representative
Non-probability samples data from a group. As In quota sampling the
opposed to random selection of the sample is
in which population
sampling, quota sampling made by the interviewer,
subgroups are classified
requires that representative who has been given
on the basis of individuals are chosen out quotas to fill from
researcher judgment.. of a specific subgroup specified sub-groups of
the population.

www.presentationgo.com
www.presentationgo.com
Networking (Snowballing)

The image part with relationship ID The image part with relationship ID The image part with relationship ID The image part with relationship ID
rId3 was not found in the file. rId5 was not found in the file. rId7 was not found in the file. rId9 was not found in the file.

Snowball sampling It may be It lowers search costs;

extremely This technique however, it introduces
is a special
nonprobability difficult or cost relies on referrals bias because the
method used when prohibitive to from initial subjects technique itself
the desired sample locate to generate reduces the likelihood
characteristic is respondents in additional subjects. that the sample will
rare. these situations. . . represent a good cross
section from the
population.

www.presentationgo.com
 Person 1
RESEARCHER RESEARCHER HAS 3
CONTACTS

Friend/contact 1 Friend/contact 2 Friend/contact 3

contacts his/her contacts his/her contacts his/her
own own own
friends/contacts/ friends/contacts/ friends/contacts/

4 5 6 7 8 9 10 11 12

THE 3 CONTACTS EACH HAVE 3 CONTACTS

www.presentationgo.com
 Determining Sample Size

Sample size can be determined based on:

• Population size (classic and traditional
way)
• Structure of the research model
(theoretical framework)

01 02 03 04 05 06 07

www.presentationgo.com
www.presentationgo.com
www.presentationgo.com
www.presentationgo.com
Sampling Technique vs Research method

www.presentationgo.com
Data collection Method

www.presentationgo.com
www.presentationgo.com
Data Collection vs Research method

www.presentationgo.com