Proportional Simple Random Sampling
Also known as: proportional SRS, probability-proportional simple random sampling, proportional random sampling
Proportional simple random sampling is a probability-based sampling technique in which units are drawn at random from each subgroup of the population in numbers proportional to each subgroup's share of the total population. This ensures the resulting sample mirrors the population's composition across key subgroups, while retaining the randomness and unbiasedness of simple random sampling within each group.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use proportional simple random sampling when you need a sample that accurately reflects the population's composition across one or more known subgroups and you want to preserve the simplicity and unbiasedness of random selection. It is especially appropriate when subgroup sizes differ considerably — so that purely unrestricted SRS could by chance under-represent small groups — and when subgroup membership is known for all population members. It is not appropriate when some subgroups are so small that proportional allocation would yield too few observations from them for reliable subgroup-level analysis (use disproportional stratified sampling instead), or when no complete sampling frame exists.
Strengths & limitations
- Guarantees that the sample's subgroup composition mirrors the population, improving face validity and reducing the risk of chance imbalances from unrestricted SRS.
- Produces unbiased estimates of population parameters under well-understood probability theory.
- Simple to implement and explain to stakeholders when the population subgroup sizes are known.
- Estimates typically have lower variance than unrestricted SRS for population means when the variable of interest varies across subgroups.
- No post-hoc weighting is required because proportionality is built into the allocation.
- Requires a complete and accurate sampling frame with known subgroup membership for every population unit — often unavailable in practice.
- Proportional allocation guarantees representation of the chosen subgroups but not of any other characteristic; the sample may still be imbalanced on uncontrolled variables.
- For very small subgroups, proportional allocation may yield sample sizes too small for reliable subgroup-specific estimates.
- Offers no efficiency gain over unrestricted SRS when the variable of interest is unrelated to the subgroup variable.
Frequently asked
Is proportional simple random sampling the same as stratified random sampling?
Structurally yes — proportional SRS is a stratified design in which random sampling is applied within each stratum and the allocation is proportional. The distinction is mostly terminological: 'stratified sampling with proportional allocation' is the more precise label found in most statistics texts. When analysts say 'proportional SRS' they almost always mean proportionally allocated stratified simple random sampling.
When should I use disproportional instead of proportional allocation?
Use disproportional (unequal) allocation when you need reliable estimates for subgroups that are small relative to the population. Proportional allocation would give those small subgroups too few observations for stable estimates. Disproportional allocation oversamples small subgroups and then applies weights in analysis to restore population-level unbiasedness.
Does proportional allocation always give more precise estimates than unrestricted SRS?
Not always. Proportional stratified SRS is at worst as precise as unrestricted SRS (the variance is never larger), and it outperforms unrestricted SRS when the variable of interest differs systematically across strata. If there is no systematic difference between strata on the outcome variable, the gains are negligible.
How do I analyze data from proportional simple random sampling?
Because the design is stratified, use stratified estimators for means, totals, and proportions, and stratified variance formulas (not the simple SRS formula). Statistical software packages such as R (survey package), Stata (svy commands), and SPSS (Complex Samples module) support stratified sampling designs and will compute correct standard errors.
What if my calculated n_h values are not whole numbers?
Round each n_h to the nearest integer, then check that the sum equals the target total n. If the sum is off by one or two due to rounding, adjust the largest or smallest strata by one unit. Document all rounding decisions in your methods section.
Sources
- Cochran, W. G. (1977). Sampling Techniques (3rd ed.). John Wiley & Sons. ISBN: 978-0471162407
- Lohr, S. L. (2009). Sampling: Design and Analysis (2nd ed.). Brooks/Cole. ISBN: 978-0495105275
How to cite this page
ScholarGate. (2026, June 3). Proportional Simple Random Sampling. ScholarGate. https://scholargate.app/en/survey-methodology/proportional-simple-random-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
- Proportional Stratified SamplingSurvey Methodology↔ compare
- Simple random samplingSurvey Methodology↔ compare
- Stratified SamplingSurvey Methodology↔ compare
- Systematic SamplingSurvey Methodology↔ compare
- Weighted SamplingSurvey Methodology↔ compare