Only a random sample of a population supports generalizing to that population.
A result of 40% with a 5-point margin of error → the true population value is plausibly between 35% and 45%.
Step-by-step strategy
Ask: who could be chosen? Does the sample represent the whole population?
Common traps
Volunteers or specific groups (a club, a library) give biased samples.
The margin of error is about the population, not the sample.
Worked example
Random survey: 52%±3% → it is plausible that between 49% and 55% of the whole population holds that view.
Sample questions
Try to answer before opening the solution.
Question 1
Difficulty: easy≈ 1 min
In a random sample of 250 voters in a town, 140 said they plan to vote in the next election. If the town has 12,000 voters, approximately how many of them plan to vote?
This is an estimate; the true number is likely close to 6,720.
Question 2
Difficulty: medium≈ 1 min
A town has 8,000 households. In a random sample of 250 of these households, 45 have a dog. Based on the sample, what is the best estimate of the number of households in the town that have a dog?
Numeric answer (you type the number).
Show answer and explanation
Answer: 1440
Explanation
In the sample, 25045=0.18, or 18%, of the households have a dog. Because the sample is random, estimate the same proportion for the town: 0.18⋅8,000=1,440 households.
Question 3
Difficulty: hard≈ 1 min
A researcher wants to reduce the margin of error of an estimate of a population proportion by half, keeping the same confidence level. If the original random sample had 400 people, approximately how many people should the new random sample have?
Numeric answer (you type the number).
Show answer and explanation
Answer: 1600
Explanation
Margin of error ∝n1. To halve the margin, n must double, so n must be multiplied by 4: 400×4=1,600.
Doubling the sample to 800 only reduces the margin by a factor of 2≈1.41.