الرياضيات
Margin of Error: Turning One Survey Number into a Range
A margin of error converts a survey result into a plausible range for the whole population. What widens it, what narrows it, and how the SAT asks about it.
Margin of error
The margin of error is the amount added to and subtracted from a survey result to give the range in which the true population value plausibly lies.
- أين يقابله الطلاب
- Digital SAT Math, in the Problem-Solving and Data Analysis domain, under the testing point named "Inference from sample statistics and margin of error".
الإجابة باختصار
A margin of error turns a single survey result into a range. If 52% of a sample answer yes and the margin of error is 3 percentage points, the plausible value for the whole population is between 49% and 55%. A larger sample narrows that range; a biased sample is not helped by it at all.
مثال
52% ± 3% → 49% to 55%
A poll of 600 registered voters finds 52% support a proposal, with a reported margin of error of 3 percentage points. The credited conclusion is that it is plausible the proportion among all registered voters is between 49% and 55%. Note what cannot be concluded: not that 52% is the answer, and not that support is above half, because 49% sits inside the interval too.
The interval is the result, not the single number
A survey measures a sample and is used to talk about a population. The sample proportion is the best single estimate available, but on its own it overstates what the survey knows. The margin of error is the correction: subtract it and add it, and you have the range the survey actually supports.
Two details of wording repay attention. The margin is quoted in percentage points, not per cent, so 3 percentage points off 52% gives 49% and not 50.44%. And the range is described as plausible rather than certain — the true population value can sit outside it, just less often than inside.
What makes the range wider or narrower
Sample size is the lever these questions test. Ask more people and the margin of error shrinks, but not in proportion: for a percentage it falls with the square root of the sample size, so cutting the margin in half takes four times as many people. That is why national polls stall around a thousand respondents rather than pushing on to ten thousand.
Confidence level pushes the other way. Demanding more certainty that the interval contains the true value makes the interval wider, so a margin of error is only meaningful alongside the confidence level it was calculated at.
What the margin of error says nothing about is whether the sample was any good. It measures the luck of the draw and nothing else. A survey of students at one university has a perfectly respectable margin of error and still tells you nothing about students anywhere else, and an online poll answered only by people who felt strongly can be precise and wrong at the same time.
How the SAT asks about it
The margin of error is given to you. There is no formula for it on the reference sheet and no question requires you to derive one, which means the whole task is interpretation. Expect to be asked for the most appropriate conclusion from a stated result, or which change to a study would reduce its margin of error, or which population a result may fairly be applied to.
The wrong answers are predictable. They state the sample figure as fact, extend the conclusion to a wider population than the one sampled, or treat the interval as a guarantee. Reading the first sentence of the item carefully — who exactly was surveyed, and who the question is now asking about — decides most of these before any arithmetic starts.
- Increasing the sample size reduces the margin of error
- Raising the confidence level increases it
- Conclusions apply only to the population the sample was drawn from
- A small margin of error does not correct a badly chosen sample
- The increase in sample size needed to halve a margin of error
- 4×The increase in sample size needed to halve a margin of error
أسئلة شائعة
What does a margin of error of 3% mean?
It means the survey result should be read as a range three percentage points either side of the figure reported. A result of 52% becomes 49% to 55%. The margin describes how much the estimate could move because a sample was measured instead of the whole population, and it applies to the reported percentage, not to each individual answer.
Does a bigger sample always reduce the margin of error?
All else being equal, yes, but with sharply diminishing returns. The margin falls with the square root of the sample size, so quadrupling the number of people surveyed roughly halves it. Going from 1,000 to 2,000 respondents buys much less improvement than the extra cost suggests, which is why most published polls sit near a thousand.
Do you have to calculate the margin of error on the SAT?
No. SAT items supply the margin of error and test whether you can use it. There is no margin of error formula on the reference sheet, and the questions ask for the appropriate conclusion, the effect of changing the sample, or the population the result applies to. Time spent memorising a formula for it is time wasted.
Can a survey with a small margin of error still be wrong?
Easily. Margin of error only accounts for the randomness of sampling. It cannot detect a leading question, a sample that excludes people without phones, or respondents who do not answer honestly. Those problems are called bias, they do not shrink when the sample grows, and no reported interval will warn you about them.
Is the margin of error the same as a confidence interval?
They are two ways of saying one thing. The confidence interval is the range itself, 49% to 55%; the margin of error is the half-width used to build it, 3 percentage points either side of 52%. Both depend on the confidence level chosen, which is why a margin of error quoted with no confidence level attached is incomplete.
المصادر
آخر تحديث
معرفة الكلمة ليست كاستخدامها
يستطيع المعلّم أن يرى الطالب وهو يستعمله في سؤال، فيتبيّن أين يتوقف الفهم بالضبط. الحصة الأولى مجانية.