Skip to contentScholarGate
LibraryBookshelfDeskReview StudioAssistant
Sign in
On this page
IntuitionHow it worksWhen to use itStrengths & limitationsCommon pitfallsApplicationsFrequently asked🔒 Read the full methodSourcesRelated methods
Cite this pageSpotted an issue on this page? Report or suggest a fix →
Home›Survey Methodology›Disproportional Stratified Sampling — Unequal-Allocation Stratified Design
Process / pipelineSampling

Disproportional Stratified Sampling — Unequal-Allocation Stratified Design

Disproportional Stratified Random Sampling · Also known as: disproportionate stratified sampling, unequal-probability stratified sampling, oversampling stratified design, non-proportional stratified sampling

Disproportional stratified sampling divides the population into mutually exclusive strata and deliberately draws different proportions from each stratum — oversampling small or analytically important subgroups and undersampling large ones. Post-hoc weighting restores population-level representativeness when overall estimates are needed. First formalised by Jerzy Neyman in 1934, it is the standard approach when subgroup-level precision matters as much as total-population estimates.

ScholarGate
  1. Process / pipeline
  2. v1
  3. 2 Sources
  4. PUBLISHED
Cite this page →
Tools & resources
Download slides
Learn & explore

Read the full method

Members only

Sign in with a free account to read this section.

Sign in

Method map

The neighbourhood of related methods — select a node to explore.

Disproportional Stratified Sampling
Cluster SamplingMultistage SamplingProportional Stratified…Simple random samplingStratified SamplingWeighted SamplingAdaptive Stratified Samp…Disproportional cluster…Weighted Stratified Samp…

When to use it

Use disproportional stratified sampling when subgroup-level estimates must meet a minimum precision threshold that proportional allocation cannot satisfy — for example, rare subpopulations, regulatory reporting requiring disaggregated estimates, or comparative studies needing equal-sized subgroup samples. It is also appropriate when stratum-level variance is known to differ markedly and Neyman optimal allocation is feasible. Do not use it when no stratifying variable exists on the sampling frame, when population strata sizes are unknown (making weight computation impossible), or when the analysis is purely descriptive at the aggregate level and proportional allocation already provides sufficient precision.

Strengths & limitations

Strengths
  • Guarantees adequate sample sizes in small but analytically important subgroups, enabling reliable subgroup inference.
  • Neyman optimal allocation can minimise total variance for a fixed overall sample size when stratum variances differ substantially.
  • Retains full probability-sampling properties: every unit has a known selection probability, supporting design-based inference.
  • Widely accepted in official statistics, health surveys, and social research; well-supported by standard statistical software.
  • Enables cost-efficient designs when data collection costs vary by stratum, by allocating more to cheaper strata.
Limitations
  • Requires a sampling frame with reliable stratum membership information for all population units before sampling begins.
  • Produces unequal sampling fractions, making unweighted analysis misleading; analysts must consistently apply design weights.
  • Optimal allocation requires prior knowledge or estimates of within-stratum variances, which may not be available for new surveys.
  • Complex weighting can inflate variance of aggregate estimates relative to proportional allocation if oversampled strata are large.

Frequently asked

How does disproportional stratified sampling differ from proportional stratified sampling?

In proportional stratified sampling, each stratum contributes to the sample in exact proportion to its share of the population, so all sampling fractions are equal. In disproportional stratified sampling, strata are sampled at deliberately different rates — some strata are oversampled, others undersampled. Both designs divide the population into strata first; only the allocation rule differs. Disproportional allocation requires post-hoc weighting to restore representativeness at the population level.

Do I always need to weight the data after disproportional stratified sampling?

Yes, whenever you want population-level estimates. Unweighted analysis treats all sampled units as equally representative, which is incorrect when sampling fractions differ across strata. Within a single stratum, unweighted analysis is valid for within-stratum estimates; weighting is necessary only when combining across strata for population-level conclusions.

What is Neyman optimal allocation and when should I use it?

Neyman optimal allocation sets n_h proportional to N_h times the within-stratum standard deviation S_h. It minimises the variance of the population mean estimator for a fixed total sample size. Use it when you have reliable prior estimates of S_h from a pilot study or administrative data and when minimising overall variance is the primary design goal. If subgroup precision is the goal, equal or minimum-threshold allocation may be more appropriate.

Can I use standard regression software, or do I need special survey analysis procedures?

You must use survey-aware procedures — for example, Stata's svy prefix, R's survey package, SAS PROC SURVEYREG, or SPSS Complex Samples — that accept design weights and the stratified design structure. Standard OLS or logistic regression routines ignore the design, producing correct point estimates only under specific conditions but systematically incorrect standard errors, confidence intervals, and p-values.

How do I handle non-response in a disproportionally stratified sample?

Non-response is handled at the weighting stage. The base design weight w_h = N_h/n_h is multiplied by a non-response adjustment factor, typically the inverse of the observed response rate within each stratum. Calibration or post-stratification weighting can further align weighted totals to known population benchmarks, reducing non-response bias.

Sources

  1. Cochran, W. G. (1977). Sampling Techniques (3rd ed.). John Wiley & Sons. ISBN: 978-0471162407
  2. Neyman, J. (1934). On the two different aspects of the representative method: The method of stratified sampling and the method of purposive selection. Journal of the Royal Statistical Society, 97(4), 558-625. DOI: 10.2307/2342192 ↗

How to cite this page

ScholarGate. (2026, June 3). Disproportional Stratified Random Sampling. ScholarGate. https://scholargate.app/en/survey-methodology/disproportional-stratified-sampling

Related methods

Cluster SamplingMultistage SamplingProportional Stratified SamplingSimple random samplingStratified SamplingWeighted Sampling

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

  • Cluster SamplingSurvey Methodology↔ compare
  • Multistage SamplingSurvey Methodology↔ compare
  • Proportional Stratified SamplingSurvey Methodology↔ compare
  • Simple random samplingSurvey Methodology↔ compare
  • Stratified SamplingSurvey Methodology↔ compare
  • Weighted SamplingSurvey Methodology↔ compare
Compare side by side →

Referenced by

Adaptive Stratified SamplingDisproportional cluster samplingProportional Stratified SamplingWeighted Stratified Sampling

Similar methods

Weighted Stratified SamplingProportional Stratified SamplingProportional Weighted SamplingDisproportional cluster samplingProportional Simple Random SamplingStratified SamplingMulti-level Stratified SamplingField-based Stratified Sampling

Related reference concepts

Study Matching and StratificationMantel-Haenszel and Stratified AnalysisStudy Design and Sample Size PlanningStatistical Power and Sample SizeSampling Distributions and Central Limit TheoremImportance Sampling

Spotted an issue on this page? Report or suggest a fix →

ScholarGate — Disproportional Stratified Sampling (Disproportional Stratified Random Sampling). Retrieved 2026-07-21 from https://scholargate.app/en/survey-methodology/disproportional-stratified-sampling · Dataset: https://doi.org/10.5281/zenodo.20539026
Quick facts
Originator
Jerzy Neyman
Year
1934
Type
Probability sampling design
DataType
Quantitative or mixed; structured survey or administrative records data
Subfamily
Sampling
Related methods
Cluster SamplingMultistage SamplingProportional Stratified SamplingSimple random samplingStratified SamplingWeighted Sampling
ScholarGate

A content-first reference library for research methods — what each one is, how it works, and where it comes from.

Open data (CC-BY)

Explore

  • Library
  • Search the library…
  • Browse by field
  • Fields
  • Journey
  • Compare
  • Which method?

Reference

  • Subjects
  • Atlas
  • Glossary
  • Methodology
  • Philosophy

Your tools

  • Bookshelf
  • Desk
  • Chat

Company

  • About
  • Pricing
  • Contact
  • Suggest a method

Entries are compiled from published sources for reference. Verifying the accuracy and suitability of any information for your own use remains your responsibility.

© 2026 ScholarGate · A research-method reference library
  • Privacy
  • Cookies
  • Terms
  • Delete account