Populations and

Samples
Nursing 200W
Chapter 9
 Objectives
 Define population, sample and sampling.
 Distinguish between nonprobability and
probability samples and compare their
advantages and disadvantages.
 Compare sampling for quantitative and
qualitative research methods.
 Evaluate the appropriateness of the
sampling method and sample size used in
studies.
 Introduction
 Sampling - process of selecting a group of
elements (people, events, behaviors) that
represent a population.
 What is the purpose of sampling?
• Selecting a portion of the population to
represent the population. A sample is a
subset of population elements.
 Definitions
 Population – the entire set of
individuals (or elements) who meet
the sampling criteria. This is the “P”
in the PICO question.
• Target population
 Accessible population – portion of the
target population to which the
researcher has reasonable access
 Definitions
 Sample – the selected group of
people (or elements) from whom
data are collected for a study.
 Element – one individual unit of a
population
• Person (adolescent males) – in nursing
research humans are the most common
elements.
• Event (hospital admission)
• Behavior (walking 30 minutes per day)
Population (target)

Accessible

Sample
Element
 Sampling Criteria
 Criteria that specify characteristics of the
target population.
 Inclusion criteria: desirable characteristics
• Able to speak/read English
• Diagnosed with diabetes within last month
• 18 to 45 years old
 Exclusion criteria: undesirable
characteristics; will not be included
• Mental illness or cognitive dysfunction
• Age <18 years
• Unable to read or speak English
 Sampling
 Goal: representativeness of population
• Target population – the entire pop. In which a
researcher is interested.
• Accessible population – the portion of the target
pop. that is accessible to the researcher.
 Representative Sample – one that closely
approximates the population; this improves
generalizability of finding to the target population.
 Sampling bias – overrepresentation or
underrepresentation of some segment of the
population; decreases generalizability
 Variations in Sampling
 Random variation
• Expected differences in values of a
variable within a sample
• Variation around a mean
• Humans are unique and different!
 Systematic variation (bias)
• Result of sampling error
• Occurs when sample is not
representative of population
 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

 Types of Systematic Variation
 Refusal rate: percentage of subjects that
declined to participate in the study
• 80 subjects approached and 4 refused
• 4  80 = 0.05 (5% refusal rate)
 Acceptance rate: percentage of subjects
that consented to be in the study
• 80 subjects approached and 76 accepted
• 76  80 = 0.95 (95% acceptance rate)
 Sampling Mortality
 Loss of subjects from the study
• Can be from one group (experimental or
control)
• One group may be more likely to
withdraw than the other.
 Can result in bias/systematic
variation
 Mortality should be reported by
researchers
 Sampling Plan – Quantitative
Studies
 Strategies used to obtain a sample
for a study
 Types of sampling plans
1. Probability
2. Non-Probability
 1. Probability Sampling
 Random selection of elements from a
population
 Selection is up to chance
 Types
• Simple random
• Stratified random
• Cluster random
• Systematic
 Simple Random Sampling
 Most basic method
 Each element in target population
has equal chance of being in the
sample
• Improves representativeness of sample
 Examples
• Draw names
• Random number table
• Computer program
 Stratified Random Sampling
 Ensures all levels of identified
variables are adequately represented
in the sample
 Need a large population to start with
 Variables often stratified
• Age, gender, socioeconomic status
• Types of nurses, sites of care
 Systematic Sampling
 Select every Kth individual from a list,
using a randomly selected starting
point
• e.g. every 10th person from a list
 Researcher must know number of
elements in the population and the
sample size desired
 Point of Clarification
 Random sampling is NOT the same
as random assignment
 In random sampling, each element in
the population has an equal
independent chance of being
selected.
 Random assignment – assigning
members of sample to experimental
or control group as seen in a
randomized control trial.
 Sampling – Key Points
 Probability sampling is the only
method for obtaining a highly
representative sample
• Reduces bias
• Improves ability to generalize results of
quantitative studies
 2. Non-probability Sampling
 Not every element of population has
opportunity for selection
• Decreases representativeness
 Reasons commonly used?
• Cost or time savings
 Types
• Convenience
• Quota
 Convenience Sampling
 “Accidental” sampling
 Whomever is available and willing to
give consent
 Weakest approach – high risk of
bias!
 Quasi-experimental studies
• use convenience sample
• subjects are randomly assigned to
groups
 Quota Sampling
 Similar to stratified random sampling
 Deliberately attempt to include
subjects from sub-groups who may
be under-represented in the
convenience sample
 Goal: replicate the proportions of
subgroups present in the population
 Better than convenience sampling
(less bias)
 Sampling Methods: Qualitative
Studies
 Representativeness of sample less
important
 Goal is to uncover meanings, not to
generalize findings
 Methods
1. Purposive
2. Networking
3. Theoretical
 1. Purposive
 Conscious selection of sample by
researcher
 Selection based on specific
characteristics
 Weakness – no way to evaluate bias
on part of the researcher
 2. Networking (Snowballing)
 Use social networks to connect with
poorly accessible populations
 Examples:
• Alcoholics
• Homeless persons
• Breastfeeding moms
 Ethical concerns
 3. Theoretical Sampling
 Used in grounded theory studies
 Selection of participants based upon
their knowledge of or experience
with the phenomenon of interest
 Sample not pre-determined
• Characteristics may change as data
collection evolves
 Sample Size
 In general – a larger sample is more
representative of the population

Does size really matter?
• Inadequate sample size may influence
ability to detect statistically significant
differences between groups – this is
important in quantitative studies.
 Power Analysis
 Definition: statistical procedure used
to estimate sample size needed to
correctly reject a null hypothesis
 Usually done prior to sampling
• May be done after data analysis, if
unable to reject null hypothesis
 Factors That Influence
Adequacy of Sample Size
1. Effect size
2. Type of research design
3. Number of variables
4. Sensitivity of measurement tools
5. Data analysis technique
 Effect Size
 Strength of relationship between variables
or magnitude of difference between groups
 Detecting a large effect size (strong
relationship or big difference)
• Easier
• Smaller sample needed
 Detecting a small effect size
• More difficult
• Larger sample needed
 Effect Size
 Increasing the sample size also
increases the effect size
 More likely to detect differences
between groups with larger sample
 Small effect <0.3
 Medium effect 0.5
 Large effect >0.6
 Sample Types in Research
Designs
 Descriptive
• Questionnaires
• Large samples needed
 Correlational
• Studies with more variables need larger
samples
 Experimental/Quasi-experimental
• More control
• Smaller samples (groups)
 Two Issues to Consider when
Critiquing a Sampling Plan
 1. Consider whether the researcher
has adequately described the
sampling strategy.
 2. Consider whether the researcher
made good sampling decisions.

This can be critiqued by considering

the following…
Critiquing Sampling (Quantitative)
 Type of sampling
 Sample size
• Rationale for sample size (power
analysis)
 Inclusion & exclusion criteria
 Method of recruiting
 Description of sample characteristics
 Refusal & sample mortality
Critiquing Sampling (Qualitative)
 Inclusion & exclusion criteria
 Method of recruiting
 Refusal and sample mortality
 Characteristics of sample
 “Quality” of participants – articulate,
well-informed, willing to share
 Sample size determined by
saturation