Interpretation: we are 95% confident between 50.1% and 59.9% of voters support the candidate
Margin of error quantifies polling uncertainty. 'Candidate leads 55% ± 5%' means the true support could be anywhere from 50% to 60%. A margin of error that doesn't include 50% still suggests the candidate leads, but we can't be certain.
Example 2
medium
A news report claims: 'The margin of error for this poll is ±3%.' Explain three things a careful reader should know about this margin of error.
Example 3
medium
You need E≤3 for a 95% CI on a mean with σ=20. Find the smallest n.
Example 4
hard
Show that p^(1−p^) is maximized when p^=0.5, and find its max value.
Practice Problems
Try each one, then open the solution to compare your method.
Example 1
easy
A 95% CI for average commute time is (28,36) minutes. What is the margin of error, and what does it tell us about the survey's precision?
Example 2
hard
Two polls: Poll A (n=100): margin of error ±10%. Poll B (n=1600): margin of error ±2.5%. By what factor was n increased, and by what factor did E decrease? Verify using E=z∗n0.5×0.5.
Example 3
easy
A poll reports 52% approval with a margin of error of ±3%. What interval of values does this imply?
Example 4
easy
A confidence interval is (40,60). What is the margin of error?
Example 5
easy
The margin of error for a mean is z∗⋅SE. If z∗=1.96 and SE =2.5, find the margin.
Example 6
easy
True or false: a larger sample size produces a smaller margin of error (same confidence).
Example 7
easy
Does the margin of error account for non-sampling bias such as bad question wording?
Example 8
easy
To reduce the margin of error, should you increase or decrease the sample size?
Example 9
easy
A 95% margin uses z∗=1.96; a 90% margin uses z∗=1.645. With the same data, which margin is larger?
Example 10
easy
A proportion estimate is p^=0.6 with margin 0.04. State the confidence interval.
Example 11
medium
A sample has σ=20, n=100. Find the 95% margin of error (z∗=1.96).
Example 12
medium
A poll has p^=0.5, n=1000. Find the 95% margin (z∗=1.96) using SE =p^(1−p^)/n.
Example 13
medium
Currently margin =6 with n=100. Keeping confidence fixed, what margin results from n=400?
Example 14
medium
Two candidates poll at 48% and 52%, each with margin ±3%. Can we declare a clear winner?
Example 15
medium
A margin of 4.9 comes from z∗=1.96 and SE =2.5. If confidence rises to 99% (z∗=2.576), what is the new margin?
Example 16
medium
A study needs a margin of at most 2 for a mean with σ=10 at 95% (z∗=1.96). Find the smallest n.
Example 17
medium
A margin of error is z∗⋅σ/n. With σ=18, n=36, z∗=1.645 (90%), find the margin.
Example 18
challenge
Quadrupling the sample size from n to 4n changes the margin by what factor, and why is this 'diminishing returns'?
Example 19
challenge
A poll uses p^=0.5 for a conservative margin. Why does p^=0.5 maximize the margin, and what is the maximum value of p^(1−p^)?
Example 20
challenge
A margin of 4 at n=50 must be cut to 1 at the same confidence and σ. What sample size is required?
Example 21
medium
A sample has σ=24, n=64. Find the 95% margin of error (z∗=1.96).
Example 22
medium
A poll of n=625 has p^=0.4. Find the 95% margin (z∗=1.96) using SE =p^(1−p^)/n.
Example 23
easy
A 95% CI for a mean is (12,18). What is the margin of error?
Example 24
easy
A 95% CI is (45,55). State the point estimate and margin of error.
Example 25
easy
With σ=15, n=225, z∗=1.96, find the margin of error.
Example 26
easy
To shrink the margin of error, which of these helps: bigger sample, smaller sample, or higher confidence?
Example 27
easy
A poll reports p^=0.62 with margin ±0.03. Give the 95% CI.
Example 28
medium
Find the 95% margin of error for p^=0.4, n=100 (z∗=1.96).
Example 29
medium
With p^=0.5, n=400, z∗=1.96, find the 95% margin of error.
Example 30
medium
With σ=30, n=100, z∗=1.96, compute the 95% margin of error.
Example 31
medium
Currently E=4 at n=50. Find E at n=200 (same z∗ and σ).
Example 32
medium
A poll has p^=0.3, n=100, z∗=1.96. Find the margin.
Example 33
medium
Currently E=4.9% at n=400, p^=0.5, 95% confidence. What is the new margin at n=1600?
Example 34
hard
Two candidates poll at 51% and 49%, each with margin ±3%. Can you declare a winner?
Example 35
hard
To cut the margin by a factor of 3, by what factor must n grow?
Example 36
hard
For a 99% margin with σ=10, E≤2, find the minimum n (z∗=2.576).
Example 37
hard
A CI is (0.30,0.42). What is the margin of error and the point estimate?
Example 38
hard
With σ=16, n=64, find the 95% margin (z∗=1.96).
Example 39
hard
A study triples its sample size from n=100 to n=300. By what factor does margin change?
Example 40
hard
A poll has p^=0.5, n=2500, z∗=1.96. Margin?
Example 41
challenge
A pollster wants E≤0.02 at 95% confidence with no prior estimate of p. Find the smallest n (z∗=1.96).
Example 42
challenge
A 95% CI for a mean is (xˉ−4,xˉ+4). If you re-run at 99% confidence with the same data, find the new margin (z95∗=1.96, z99∗=2.576).
What to focus on
Core idea:The margin of error is half the width of a confidence interval — how far the estimate might be from the truth.
Common stuck point:The procedure for margin of error is the easy part; the trap is reporting the full interval width as the margin of error. Asking "Am I reporting how far an estimate might be from the truth as a single ± value (half the interval width)?" first is what keeps a correct-looking calculation from being attached to the wrong concept.
Ask yourself:Am I reporting how far an estimate might be from the truth as a single ± value (half the interval width)?
Concept recap
The maximum expected difference between the sample statistic and the true population parameter; it is half the width of a confidence interval.
The tolerance on a machined bolt, "50 mm ± 0.2 mm," works like a poll's "± 3 points": a half-width on either side of a center. The difference is that a poll's margin holds only at its confidence level, about 95% of the time, and it says nothing about bias in how the sample was collected.