Stratified Sampling
Dividing a population into groups and sampling proportionately from each.
Short answer
What is Stratified Sampling?
Stratified sampling divides the population into subgroups — by booth, region, community or demographic — and draws a random sample from each in proportion to its size, ensuring every relevant group is represented.
Definition
Stratified sampling divides the population into subgroups — by booth, region, community or demographic — and draws a random sample from each in proportion to its size, ensuring every relevant group is represented.
Why It Matters
Convenience sampling produces confident numbers that describe nobody. Stratification is what makes a constituency survey capable of supporting segment-level analysis rather than only a topline figure.
Example in Indian Elections
A constituency survey stratified across booth clusters and weighted for turnout propensity supports statements about specific segments; an unstratified street-corner sample supports none.
Related Terms
- Vote ShareThe percentage of total valid votes polled that a candidate or party receives.
- SwingThe change in vote share between two elections, measuring movement of support.
- Turnout PropensityThe modelled likelihood that a given voter will actually vote.
- Cross-TabulationBreaking survey results down by demographic or behavioural subgroups.
- Sampling Error / Margin of ErrorThe uncertainty inherent in estimating a population from a sample.
- Back-CheckRe-contacting a share of survey respondents to verify enumerator honesty.