How it works
The margin of error is the “plus or minus” reported with a poll. A result of 52% with a ±3.1% margin at 95% confidence means that if the poll were repeated many times, about 95% of the resulting intervals would contain the true value.
The margin shrinks with the square root of the sample size, so quadrupling the sample halves the margin. It is widest when the proportion is 50%, which is why pollsters often quote that worst case.
The margin of error only covers random sampling error. It does not account for biased question wording, non-response or a sample that does not represent the population.
Formula
MOE = z × √( p(1 − p) ÷ n ) × √( (N − n) ÷ (N − 1) )
Questions people ask
What is the margin of error for 1,000 respondents?
About ±3.1% at 95% confidence when the proportion is near 50%. At 99% confidence it rises to about ±4.1%.
Does the margin of error apply to subgroups?
Each subgroup has its own, larger margin, because it has fewer respondents. A subgroup of 250 from a 1,000-person poll has a margin of about ±6.2%.
If two candidates are within the margin of error, is it a tie?
Not exactly. The margin on the gap between two candidates is roughly double the reported margin, so a lead smaller than that is not statistically clear, but the leader is still more likely ahead.