SAMPLING

BY: KACHIRI T. SALIBIO-MERCADAL

OBJECTIVES

• Define sampling and its importance to research

• Identify and describe several types of sampling designs.

• Evaluate the appropriateness of the sampling method and sample

size used in a study

SAMPLING

• Involves selecting
people, events,
behaviors, or other
elements with which
to conduct a study.
POPULATION

• Is the entire group of interest

• TARGET POPULATION is the
entire population of interest.
• ACCESSIBLE POPULATION is
the portion of the target population
accessible to the researcher.
• Is specified by the researcher using
ELIGIBILITY CRITERIA
 POPULATION

TARGET POP’N

ACCESSIBLE POP’N

SAMPLE

ELEMENTS
POPULATION STUDIES

The entire population is the Use data available in large

target of the study databases
 • Assumes the presence of a
HYPOTHETICAL population that can not be defined
POPULATION by a list of members of all the
population
 Include a list of characteristics essential for
membership or eligibility in the target
population.
SAMPLING OR Based from the research problem, purpose,
ELIGIBILITY RRL, conceptual and operational definitions
CRITERIA of study variables and the design.

Heterogenous vs Homogenous samples

INCLUSION EXCLUSION
CRITERIA/SAMPLING CRITERIA/SAMPLING
• Characteristics that a • Characteristics that can
subject or element must cause a person or element
possess to be part of the to be eliminated or
target population excluded from the target
population.
 SAMPLE REPRESENTATIVENESS

CHARACTERISTICS OF SAMPLES IS USUALLY EVALUATED BY

MUST BE SIMILAR TO THE TARGET TO ATTAIN GENERALIZING COMPARING THE NUMERICAL
POPULATION IN AS MANY WAYS AS VALUE OF THE SAMPLE (EX. MEAN)
POSSIBLE. WITH THE SAME VALUES FROM THE
TARGET POPULATION
(POPULATION) PARAMETER

• A numerical value of a population.

• Can be estimated from the identified values obtained in previous studies examining the
same variables.
• PRECISION is the accuracy with which the population parameters have been estimated
within a study
SAMPLING ERROR
 LARGE sampling error
Is the difference between a
means that the sample
sample statistic and the
statistic does not provide a Inversely proportional
estimated population
precise estimate of the with sample size.
parameter that is actual
population parameter; It is
but unknown.
NOT REPRESENTATIVE

Results from either

Reduces the POWER of
Random Variation or
the study
Systemic Variation

SAMPLING ERROR
 • Is the normally-occurring and expected difference
in values that occurs when one examines different
RANDOM subjects from the same sample.
VARIATION • Values of individual subjects vary from the value of
the sample mean.
• The difference is random because the value of
each subject is likely to vary in value and direction
from the previously-measured one.
• As the sample size becomes larger, overall
variation in sample values decreases, with more
values being close to the sample mean.
 • Bias or variation obtained when subjects in a
study share various characteristics, making the
sample less representative than desired. Their
resemblance to one another makes it more
SYSTEMATIC BIAS likely that demographics and measurements of
effects of interventions will be quite similar for
OR VARIATION most of them.
• Sampling error is related to the nonrandom
sampling process.
• Increases as the REFUSAL Rate increases.
REFUSAL RATE

• Is the number & percentage of subjects who decline to participate in the study.
• The higher the refusal rate, the less representative the sample is of the target population.
• Formula:
Refusal Rate = number of potential subjects refusing to participate
--------------------------------------------------------------------------
Number of potential subjects meeting sample criteria
x 100%
ACCEPTANCE RATE

• The number and percentage of the subjects who agree to participate in a study.

Acceptance Rate = number of potential subjects agreeing to participate

--------------------------------------------------------------------------
Number of potential subjects meeting sample criteria
x 100%
SAMPLE ATTRITION RATES

• The withdrawal or loss of subjects or study participants from a study before its
completion.
• Researchers must publish the attrition rate and the reason behind
• Low attrition rate allowable range: <10% - 15%

Attrition Rate = (number of subjects withdrawing)

------------------------------------------------- x 100%
Sample size
RETENTION RATE

• Opposite of attrition rate

• The number and percentage of subjects completing the study
• The higher the retention rate, the more representative the sample is of the target
population, and the more likely the study results are an accurate reflection of reality.

Retention Rate = (number of subjects withdrawing)

------------------------------------------------- x 100%
Sample size
 In random variation

In systematic variation
IN
SUMMARY….SAMPLING When sampling is not random
ERROR CAN OCCUR
High refusal rates/ low acceptance rate

High attrition rates/ low retention rate

In small sample sizes

RANDOMIZATION
SAMPLING FRAME
SAMPLING PLAN

Describes the strategies that will be used to obtain a sample for a study.

To enhance representativeness

Reduce systematic bias

Decrease sampling error

 The detailed process of selecting a group
of people, events, behaviors, or other
elements that represent the population
being studied.

SAMPLING
METHOD Quantitative Designs use Probability and
Nonprobability Sampling designs

Qualitative Designs and Mixed Methods

Research use Nonprobability Sampling
designs.
PROBABILITY (RANDOM)
SAMPLING METHOD

• Means that every member (element) of the

population has a greater than zero
opportunity to be selected for the sample
• Aka random sampling methods
• More likely represent the population
and decrease sampling error
• Thus validity of the study increase
 Simple Random Sampling
4 TYPES OF Stratified Random Sampling
PROBABILITY
SAMPLING
DESIGNS:
Cluster Sampling

Systematic Sampling
SIMPLE
RANDOM
SAMPLING
SIMPLE RANDOM SAMPLING WITH REPLACEMENT
STRATIFIED • Used when the researcher knows some of the variables in
RANDOM the population that are critical to achieving
SAMPLING representativeness
• Age, gender, ethnicity, SES, diagnosis, geographical region,
type of institution, type of care, care provider, and site of
care
• 2 types:
• A. Disproportionate sampling (equal number per strata)
• B. Proportionate sampling (selected in proportion to their
occurrence in the population
STRATIFIED RANDOM SAMPLING
• Is similar to stratified random sampling but takes CLUSTER
advantage of the natural clusters or groups of
SAMPLING
population units that have similar characteristics.
• 1. a simple random sampling would be prohibitive in
terms of time and cost
• 2. if individual elements making up the population are
unknown, preventing the development of a sampling
frame.
• Multistage cluster sampling
• Advantage: can get a larger sample size at a
lower cost than simple random sampling
CLUSTER SAMPLING
SYSTEMATIC SAMPLING

• Can be conducted when an ordered list of the population is available.

• The process involves selecting every kth individual in the list
• Starting point must be RANDOMLY SELECTED.

k = population size / sample size desired

NONPROBABILITY
(NONRANDOM) SAMPLING
FOR QUANTI & QUALI
RESEARCH
 Not every element of the
population has an opportunity to be
included in the sample

NONPROBABILITY
(NONRANDOM) Increase the likelihood of obtaining
samples that are not representative
SAMPLING of their target populations.
METHOD

Researchers often include any

subjects willing to participate who
meet the eligibility criteria
TYPES OF PURPOSIVE SAMPLING

Maximum
Homogeneous Typical case Extreme case
variation
sampling sampling sampling
sampling

Stratified
Intensity Reputational
purposeful
sampling case sampling
sampling
SAMPLE SIZE IN
QUANTITATIVE
RESEARCH
 The deciding factor in determining an adequate
sample size for correlational, quasi-experimental,
and experimental studies is POWER

POWER
POWER is the capacity of the study to detect
differences or relationships that actually exist in
the phenomenon.

0.80 or 80% minimum acceptable power

POWER ANALYSIS

Usually include:
The method in
standard power of 0.8,
determining adequate
level of significance set
sample size in
at 0.05, effect size, and
quantitative research
sample size.
1. The more stringent the significance level (ex. 0.001 vs
0.05), the greater the necessary sample size. KRAEMER &
2. Two-tailed statistical tests require larger sample than THIERMAN
one-tailed tests. (1987)
3. The smaller the effect size (ES), the larger the needed IDENTIFIED
sample size. FACTORS TO
4. The larger the power required, the larger the needed CONSIDER IN
sample size. DETERMINING
5. The smaller the sample size, the smaller the power of SAMPLE SIZE
the study
6. Factors that affect power: ES, type of study, number of
variables, sensitivity of measurement methods and
data analysis techniques.
SAMPLE SIZE IN
QUALITATIVE STUDIES
The sample size and sampling plan are determined by the
purpose and philosophical basis of the study.

Saturation of data aka Occurs when additional sampling provides

no new information, only redundancy of
Informational Redundancy previously collected data.
 RECRUITING &
RETAINING RESEARCH
PARTICIPANTS
NEXT STEP AFTER SAMPLING PLAN DEVELOPMENT
 • Involves identifying, accessing, and
RECRUITING
communicating with potential study
RESEARCH
participants who are representative of the
PARTICIPANTS
target population.
RECRUITMENT

Recruitment
Courtesy Persistence
method

Sharing
Incentives Benefits
results

Convenience Endorsements
 INVOLVES THE
PARTICIPANTS RETAINING
COMPLETING THE
REQUIRED BEHAVIORS OF RESEARCH
A STUDY TO ITS
CONCLUSION.
PARTICIPANTS