Sampling Guide- Study Guide
Overview
This study guide provides a comprehensive overview of sampling methodologies, covering
fundamental concepts, the necessity for sampling, probability and non-probability techniques,
and the development and execution of a sampling plan. It emphasizes both theoretical
understanding and practical application, equipping you with the knowledge to design and
implement effective sampling strategies.
Basic Sampling Concepts and Definitions
Key Concepts
Concept Definition
The entire group of individuals, objects, or events of interest in a
Population
study.
Sample A subset of the population selected for study.
Sampling The process of selecting a sample from a population.
Sampling Unit An individual member of the sample; the basic unit sampled.
The entire group of individuals, objects, or events of interest in a
Target Population
study.
Study (or Sampled)
The population from which the sample is actually drawn.
Population
A list of all the sampling units in the population from which the
Sampling Frame
sample will be selected.
The degree to which a sample accurately reflects the characteristics
Representativeness
of the population.
The extent to which findings from a sample can be applied to the
Generalizability
population.
The difference between sample characteristics and population
Sampling Error
characteristics due to random variation.
Concept Definition
Errors during data collection, processing, or analysis unrelated to
Non-Sampling Error
the sampling process.
A systematic error in the sampling process leading to over- or under-
Bias
representation.
Sampling Process The series of steps involved in selecting a sample.
Important Relationships
A well-defined population is the foundation of any sampling strategy.
Representativeness is influenced by the sampling procedure, sample size, and
participation rate.
Generalizability is strongest with probability-based sampling methods.
Sampling error is inherent but can be quantified and reduced.
Non-sampling error can be reduced through careful planning and quality control.
Bias can be introduced at any stage; careful attention is essential.
Critical Information
Clearly define inclusion and exclusion criteria for your population.
A biased sample cannot be generalized to the population.
Minimize both sampling and non-sampling errors for accurate results.
The sampling frame is the operationalization of the population.
Sample Size Determination is a critical step in the sampling process.
Krejcie and Morgan (1970) Sample Size Formula: A common formula for determining
sample size in quantitative research.
The Need for Sampling and Sources of Error
Key Concepts
Concept Definition
Sampling
A list of all sampling units from which a sample can be drawn.
Frame
Sampling Unit The entities on which measurements are made during a survey.
Unbiased An estimator where the mean of the sample means (expected value) equals
Estimator the true population mean, resulting in zero bias.
Important Relationships
Sampling is used because studying the entire population is often impractical or
impossible.
A well-chosen sample allows researchers to make inferences about the population.
Representativeness is crucial for generalizability.
Sampling error is inherent in sampling, but its magnitude can be estimated and
controlled.
Non-sampling errors introduce bias and can significantly distort results. These errors
should be minimized through careful study design and execution.
Increasing sample size generally reduces sampling error but does not eliminate non-
sampling errors.
Probability sampling methods are generally preferred because they allow for the
calculation of sampling error and reduce the risk of bias.
The target population should be clearly defined before sampling to ensure that the
sample accurately reflects the group of interest. The study (or sampled) population is
the portion of the target population accessible for sampling.
The sampling frame's quality directly impacts the representativeness of the sample. An
incomplete or inaccurate sampling frame can introduce bias.
Critical Information
Why Sample? Cost-effectiveness, Impracticality of Studying Entire Population, Efficiency,
Enables Precise Estimates.
Factors Influencing Sample Representativeness: Sampling procedure, Sample size,
Participation (response rate).
Sampling Error vs. Non-Sampling Error: Sampling error is random and quantifiable. Non-
sampling error is systematic (bias) and difficult to quantify.
Key Considerations When Sampling: Clearly defined objectives, Clearly defined target
population, Geographical coverage, Factors likely to influence survey results, Appropriate
sampling unit(s), Generalizability of conclusions, Feasibility within time and budget,
Practicality of the sampling procedure, Ability of the sampling scheme to address survey
objectives with appropriate precision.
Bias Example: If the true mean weight of women aged 15-44 in Ubungo is 61.7 kg, but a
survey consistently estimates it as 59.4 kg, the estimate is biased. If repeated surveys
yielded a mean of means of 62.9 kg, the bias would be 62.9 - 61.7 = 1.2 kg.
Unbiased Estimation: If the (mean of the means) = (true mean), then the bias is zero.
Probability Sampling Techniques
Key Concepts
Concept Definition
Every member of the population has a known, non-zero
Probability Sampling
chance of being selected.
Every subset of a specified size n from the population has an
Simple Random Sampling (SRS)
equal chance of being selected.
Selecting a starting point randomly and then selecting every
Systematic Sampling
kth item thereafter.
Dividing the population into subgroups (strata) and then
Stratified Sampling
randomly sampling from each stratum.
Proportionate Stratified Sampling Each stratum has the same sampling fraction.
Disproportionate Stratified
The sampling fraction varies across strata.
Sampling
Dividing the population into clusters and then randomly
Cluster Sampling
selecting clusters to sample.
Involves more than one stage of sampling, often combining
Multi-Stage Sampling
different sampling designs.
Concept Definition
Probability Proportional to Size Clusters are selected with probability proportional to their
(PPS) Sampling size.
Important Relationships
SRS and Systematic sampling form the basis for more complex sampling methods like
stratified and cluster sampling.
Stratified sampling aims to increase precision by reducing variance within strata, while
cluster sampling is used for cost-effectiveness when populations are geographically
dispersed or lack a complete sampling frame.
Proportionate and disproportionate stratification both improve precision compared to
SRS, but disproportionate is better when variances differ across strata.
Multi-stage sampling combines different sampling techniques to address complex
population structures and resource limitations.
Critical Information
Advantages of Probability Sampling: Minimal knowledge of population needed, high
external and internal validity, statistical estimation of error is possible.
Disadvantages of Probability Sampling: Can be high cost, requires a sampling frame,
and doesn't utilize researcher's expertise.
Proportionate Stratum Formula: nh = (Nh / N) * n
Disproportionate Stratum Formula: nh = n * ( Nh * Sh ) / [Σ (Ni * Si ) ]
Krejcie and Morgan (1970) Sample Size Determination Formula: Sample size= x²NP(1−P)
/ C²(N-1)+ x²P(1-P)
Non-Probability Sampling Techniques
Key Concepts
Concept Definition
Non-Probability The probability of each member of the population being selected is
Sampling unknown.
Concept Definition
Researcher selects participants based on specific characteristics
Purposive Sampling
relevant to the study.
Existing participants recruit future subjects from among their
Snowball Sampling
acquaintances.
Sample is constructed to reflect the proportions of subgroups within
Quota Sampling
the population.
Participants are selected based on their availability and willingness to
Convenience Sampling
participate.
Important Relationships
Probability sampling involves random selection, allowing for generalizations. Non-
probability sampling does not, limiting generalizability.
The effectiveness of purposive sampling relies heavily on the researcher's expertise.
Snowball sampling is particularly useful when studying populations that are hidden or
difficult to access.
Quota sampling aims to mirror population composition but doesn't guarantee overall
representativeness.
Convenience sampling is prone to selection bias.
Critical Information
Purposive Sampling: Useful for in-depth qualitative research, but limited generalizability.
Snowball Sampling: Effective for reaching hidden populations, but potential for bias.
Quota Sampling: Relatively quick and inexpensive, but selection within quotas is non-
random.
Convenience Sampling: Easy and inexpensive, but high risk of bias.
When to use Non-Probability Sampling: When random sampling is not feasible, for
exploratory research, or when resources are limited.
Limitations: Results should be interpreted with caution, as they may not accurately
reflect the larger population.
Developing and Executing a Sampling Plan
Key Concepts
Concept Definition
Data Collection The systematic process of gathering observations or measurements in
Method research.
Sample Size The number of elements selected to participate in the study.
Sampling Method The technique used to select elements from the sampling frame.
Operational
Detailed instructions for how the sampling plan will be implemented.
Procedures
Important Relationships
The target population dictates the sampling frame.
The data collection method influences the sampling method.
The objectives of the study, financial resources, time limitations, and nature of the
problem all influence the sampling method selected.
The sampling method, acceptable error, and desired confidence level influence the
sample size.
Operational procedures are essential for ensuring the sampling plan is executed
correctly and consistently.
Critical Information
Steps in Developing a Sampling Plan: Define the Target Population, Select the Data
Collection Method, Identify the Sampling Frame, Select the Sampling Method,
Determine Sample Size, Develop Operational Procedures, Execute the Sampling Plan.
Key Considerations for Sample Size Determination (Probability Samples): Acceptable
Error, Levels of Confidence.
Important Note: Non-probability samples should not be used to make inferences to a
population.
Quick Review
Population vs. Sample: Sampling allows inferences about a population from a smaller
group.
Sampling Error vs. Non-Sampling Error: Minimize both for accuracy.
Probability Sampling: Allows for statistical inferences; includes SRS, systematic,
stratified, cluster, and multi-stage sampling.
Non-Probability Sampling: Useful for exploratory research; includes convenience,
purposive, snowball, and quota sampling.
Sampling Plan: Requires defining the population, selecting a method, determining
sample size, and developing operational procedures.
Representativeness: Key to generalizability.
Bias: Avoid systematic errors in sampling.
Krejcie and Morgan (1970): A common formula for sample size determination.
