Political Dictionary
Margin of Error
The margin of error estimates the range of sampling uncertainty around a poll result.
Definition
The margin of error is a statistical estimate of how much a result from a probability sample may differ from the true value in the population because only part of the population was surveyed. It does not capture every source of error, such as biased questions, nonresponse, turnout modeling, or poor sampling frames.
Why It Matters
Understanding the margin of error prevents overinterpreting small differences between candidates or groups. A narrow lead may represent a statistical tie.
How It Works
The margin is calculated using sample size, estimated proportion, confidence level, and design effects. It often applies to the full sample; subgroup margins are usually larger.
History
Margins of error became standard in public reporting as scientific polling and statistical sampling developed.
Example
If Candidate A has 48 percent and Candidate B has 46 percent with a ±3-point margin, the poll does not establish with certainty that A leads.
Common Misconceptions
- The margin of error includes every possible polling mistake.
- A result inside the margin means the poll is useless.
- Every number in a poll has the same margin of error.
Related Terms
- Opinion Poll
- Sampling Error
- Confidence Interval
- Polling Average
- Likely Voter
- Statistical Tie
Related Topics
See Also