✦ScienexySCIENCE KNOWLEDGE LIBRARY
topic · SCI-CHE-005-016

Random, Systematic, and Stratified Sampling

Random, systematic, and stratified sampling are three important sampling designs used to select observations from a population. They differ in how sampling units are chosen and how population variation is represented.

Explanation

Random sampling selects units using a random mechanism so that selection is not deliberately determined by the sampler's preference. In simple random sampling, each eligible unit has an equal probability of selection. Systematic sampling selects units at regular intervals after an initial starting point has been chosen. It is useful for obtaining consistent spatial or temporal coverage, but it can be affected by periodic patterns in the population. Stratified sampling first divides the population into non-overlapping subgroups called strata, based on characteristics expected to influence the measured property. Samples are then selected from each stratum. This approach can improve precision when units within each stratum are relatively similar and differences between strata are important. The three designs answer different sampling problems. Random sampling is useful when a population can be listed or accessed and selection probabilities are needed. Systematic sampling is useful when the material extends across a line, area, production process, or time period. Stratified sampling is useful when the population contains identifiable groups such as soil types, production batches, or water-depth zones. The design must be chosen before collection, and the analysis must respect the selection probabilities and weighting structure.

Variables

SymbolMeaningUnit

N

Total number of population units

Number of units

n

Total number of selected sample units

Number of units

k

Systematic sampling interval

Number of population units

Nₕ

Population size of stratum h

Number of units

nₕ

Sample size allocated to stratum h

Number of units

x̄ₕ

Mean result for stratum h

Same as the measured quantity

x̄st

Overall stratified mean

Same as the measured quantity

Example or application

A worked example has not yet been added.

Definition

Random sampling selects units by a random process; systematic sampling selects units at fixed intervals; stratified sampling divides a population into groups and selects samples from each group.

Purpose

To provide structured methods for selecting sampling units while controlling selection bias and ensuring that relevant population variation is represented.

Importance

These sampling designs influence the reliability, precision, and generalizability of analytical findings. Choosing an appropriate design helps prevent underrepresentation of important population groups or locations.

Key Concepts

  • Simple random sampling: each population unit has an equal selection probability.

  • Random start: the initial position selected randomly before systematic sampling.

  • Sampling interval: the spacing between consecutive systematic selections.

  • Stratum: a defined subgroup of a population.

  • Stratified random sampling: random selection of units separately within each stratum.

  • Proportional allocation: sample sizes are assigned in proportion to stratum population sizes.

  • Disproportionate allocation: sample sizes differ from population proportions to meet analytical objectives.

  • Periodic variation: a repeating pattern that may interact with a systematic sampling interval.

  • Sampling probability: the probability that a particular unit is selected.

Principle

The design controls how population units enter the sample. Random selection provides a probability-based mechanism, systematic selection provides regular coverage, and stratification ensures that defined subgroups are represented. Correct statistical interpretation depends on the actual design, including unequal selection probabilities and any weighting required.

Working Process

  • Define the target population and establish the sampling frame or physical sampling area.

  • Identify known sources of spatial, temporal, or material variation.

  • Choose random, systematic, or stratified sampling according to the analytical objective.

  • For random sampling, select units using a suitable random-number procedure.

  • For systematic sampling, choose a random starting point and apply a defined interval.

  • For stratified sampling, define non-overlapping strata and determine how many units to select from each.

  • Collect and label the selected units without changing the selection procedure informally.

  • Analyze the results using statistical methods consistent with the sampling design.

Equations

NameExpressionDescriptionConditions

Systematic sampling interval

k = N / n

Gives the approximate interval when selecting n units from an ordered population of N units.

The population must have a defined ordering and the interval must be suitable for its structure.

Proportional allocation in stratified sampling

nₕ = n(Nₕ / N)

Allocates the total sample size among strata in proportion to their population sizes.

The stratum population sizes must be known; practical integer rounding may be required.

Weighted mean for stratified results

x̄st = Σ(Nₕx̄ₕ) / N

Combines stratum means using their population sizes as weights.

The strata must collectively cover the population and the stratum means must be appropriately estimated.

Units

  • Population and sample sizes: number of units

  • Systematic spatial interval: m or km

  • Systematic time interval: min, h, or day

  • Analyte concentration: mg/kg, mg/L, or another appropriate unit

Examples

TitleProblemSolutionAnswerLearning Point

Calculating a systematic sampling interval

A production line contains 1,000 ordered packages. A quality-control team needs to select 100 packages systematically. Determine the interval.

k = N / n = 1000 / 100 = 10. After choosing a random starting position among the first 10 packages, the team selects every 10th package.

The systematic sampling interval is 10 packages.

A random starting point helps prevent the sampler from choosing a convenient initial position.

Proportional stratified allocation

A region contains 600 units in Stratum A and 400 units in Stratum B. A total of 50 samples must be collected proportionally.

For A: nA = 50 × 600/1000 = 30. For B: nB = 50 × 400/1000 = 20.

Collect 30 samples from Stratum A and 20 from Stratum B.

Proportional allocation preserves the population proportions in the sample.

Applications

  • Random selection of laboratory samples from a production batch.

  • Systematic monitoring of river sections and groundwater wells.

  • Regular collection of samples from industrial production lines.

  • Stratified investigation of soil types and land-use zones.

  • Quality-control sampling across different product categories.

Advantages

  • Random sampling provides a clear probability-based selection mechanism.

  • Systematic sampling offers straightforward and regular coverage.

  • Stratified sampling ensures that important subgroups are represented.

  • Stratification can improve precision when within-stratum variation is lower than overall population variation.

Limitations

  • Simple random sampling may accidentally underrepresent small but important subgroups.

  • Systematic sampling may be biased when periodic population patterns align with the sampling interval.

  • Stratified sampling requires reliable information for defining the strata.

  • Disproportionate allocation requires appropriate weighting for population-level estimates.

  • A poorly defined sampling frame can exclude important population units.

Common Mistakes

  • Confusing random sampling with selecting samples casually.

  • Starting systematic sampling at a convenient location without randomization.

  • Ignoring periodic variation in systematic sampling.

  • Defining overlapping or incomplete strata.

  • Combining stratum results without appropriate weights.

  • Assuming proportional allocation is always the most statistically efficient design.

Prerequisites

  • Introduction to analytical sampling.

  • Sampling techniques and strategies.

  • Basic arithmetic, averages, and population concepts.

Learning Objectives

  • Explain random, systematic, and stratified sampling.

  • Calculate a systematic sampling interval.

  • Calculate proportional sample allocation among strata.

  • Explain why a random start is used in systematic sampling.

  • Identify situations in which each sampling design is appropriate.

  • Recognize limitations that may introduce bias.

Practice Questions

  • How does simple random sampling differ from systematic sampling?

  • A population contains 2,400 units and 120 must be sampled systematically. Calculate k.

  • Why is a random starting point important in systematic sampling?

  • A population has 700 units in one stratum and 300 in another. Allocate 80 samples proportionally.

  • What is a stratum, and why is stratification useful?

  • How can periodic variation affect systematic sampling?

Key Takeaway

Random, systematic, and stratified sampling are different probability-based designs. Their correct use depends on population structure, the sampling objective, and the statistical interpretation required.

References

TitleAuthorsTypeRelevance

Quantitative Chemical Analysis

Daniel C. Harris and Charles A. Lucy

textbook

Analytical sampling and treatment of measurement data.

Eurachem Guide: Measurement Uncertainty Arising from Sampling

technical guide

Sampling design and uncertainty.

ISO 2859 Series: Sampling Procedures for Inspection by Attributes

international standard series

Acceptance sampling principles for inspection.

Sources and verification

Sources are shown as provided in the project data. Not every source has yet been independently verified. Please report errors or missing citations.

  • Source verification pending; check an authoritative textbook, university, standards body, or peer-reviewed source.

Suggest a correction or source · More in Analytical Chemistry