Standard error (SE) is the estimated standard deviation of a sampling distribution, calculated from sample data (like SE = s/√n for a mean). It measures how much a statistic typically varies from sample to sample, and it's the building block of every margin of error and test statistic in AP Statistics.
Standard error answers one question: if you took your sample again, how different would your statistic probably be? It's the standard deviation of a sampling distribution, except you estimate it from your actual data instead of from unknown population parameters. For a sample mean, SE = s/√n, where s is the sample standard deviation (AP Stats 7.2.D). For a difference of two means, SE = √(s₁²/n₁ + s₂²/n₂). For a difference of two proportions, SE = √(p̂₁(1-p̂₁)/n₁ + p̂₂(1-p̂₂)/n₂).
Notice what all those formulas have in common. Sample size n sits in the denominator under a square root, so bigger samples shrink the standard error and make your estimate more precise. That single fact explains why confidence intervals narrow as n grows (AP Stats 7.3.C, 9.3.C) and why larger samples give tests more power (AP Stats 6.7.C). Also, swapping s in for σ is exactly why means use t-distributions instead of z. The extra uncertainty from estimating the spread puts more area in the tails (AP Stats 7.2.A).
Standard error is the thread running through the entire second half of AP Stats. It first shows up conceptually in Unit 5 as the standard deviation of a sampling distribution (Topic 5.8 gives you σ(x̄₁-x̄₂) = √(σ₁²/n₁ + σ₂²/n₂)), then becomes the workhorse of inference in Units 6, 7, and 9. Every confidence interval is statistic ± (critical value)(SE), per AP Stats 7.2.D and 6.8.C. Every test statistic is (statistic − parameter)/SE, per AP Stats 7.9.A. Even Unit 9 regression inference follows the same template, with computer output handing you the SE of the slope. The College Board explicitly says interval formulas don't need to be memorized because you can build them from the standard error formulas printed on the AP formula sheet. If you understand SE, you can reconstruct almost every inference procedure in the course. It also drives error probabilities, since AP Stats 6.7.C states that decreasing standard error decreases the probability of a Type II error.
Keep studying AP Statistics Unit 7
Sampling Distribution (Unit 5)
Standard error is just the sampling distribution's standard deviation, estimated from data. Unit 5 gives you the theoretical version with σ; Units 6-9 replace σ with s or p̂ because you never actually know the population values. Same idea, real-world version.
Margin of Error (Units 6-7)
Margin of error is critical value × standard error. SE measures the raw sample-to-sample wobble; multiplying by t* or z* converts that wobble into the cushion around your estimate at a chosen confidence level (AP Stats 7.2.D).
Degrees of Freedom and t-distributions (Unit 7)
Using s instead of σ in the SE adds uncertainty, and the t-distribution's fatter tails account for it (AP Stats 7.2.A). For two-sample means, the degrees of freedom land somewhere between the smaller of n₁-1 and n₂-1 and n₁+n₂-2 (AP Stats 7.9.A).
Type II Errors and Power (Unit 6)
AP Stats 6.7.C lists a smaller standard error as one of the things that reduces the probability of a Type II error. Less noise in your estimate makes a real effect easier to detect, which is the same as saying the test gains power.
Multiple-choice questions love handing you two-sample data and asking you to assemble the correct confidence interval, which means picking the right SE formula. Fiveable practice questions on comparing internet access across counties or exercise rates across cities all hinge on computing √(p̂₁(1-p̂₁)/n₁ + p̂₂(1-p̂₂)/n₂) correctly. On FRQs, SE appears two ways. In inference problems like the 2018 FRQ on estimating the proportion of recycling students, you use SE inside a full interval or test. In regression problems like the 2026 investigative task, computer output gives you the standard error of the slope and you have to use it to build a t-interval or interpret precision. The formula sheet provides the SE formulas, so the exam tests whether you can choose the right one, plug in correctly, and explain what the resulting precision means in context.
Standard deviation measures spread in one set of data (how far individual values fall from their mean). Standard error measures spread in a statistic across repeated samples (how far x̄ or p̂ would typically fall from the truth if you resampled). The giveaway is the √n in the denominator. SE = s/√n shrinks as the sample grows, because averaging more data smooths out the noise, while the standard deviation of individuals doesn't shrink at all. If a question asks about variability of people or measurements, that's standard deviation; if it asks about variability of an estimate, that's standard error.
Standard error is the estimated standard deviation of a sampling distribution, telling you how much a statistic like x̄ or p̂ typically varies from sample to sample.
For a sample mean, SE = s/√n; for a difference of means, SE = √(s₁²/n₁ + s₂²/n₂); for a difference of proportions, SE = √(p̂₁(1-p̂₁)/n₁ + p̂₂(1-p̂₂)/n₂).
Every confidence interval in the course follows the pattern statistic ± (critical value)(standard error), and every test statistic follows (statistic − parameter)/(standard error).
Because n is under a square root in the denominator, quadrupling the sample size only cuts the standard error in half, which is why interval width is proportional to 1/√n.
Using the sample standard deviation s instead of σ in the SE is exactly why inference for means uses t-distributions, which put more area in the tails than the normal curve.
A smaller standard error decreases the probability of a Type II error, meaning the test has more power to detect a false null hypothesis.
Standard error is the estimated standard deviation of a sampling distribution, computed from sample data. For a sample mean it's SE = s/√n, and it tells you how much your statistic would typically vary if you repeated the sampling.
No. Standard deviation describes spread among individual data values, while standard error describes spread of a statistic across repeated samples. SE always has sample size in the denominator (s/√n), so it shrinks as n grows; the standard deviation of individuals doesn't.
No. The standard error formulas are printed on the AP Statistics formula sheet. The College Board's clarifying statement says interval and test formulas can be constructed from the general formulas and the SE formulas provided, so your job is choosing and applying the right one.
It decreases, but slowly, since SE is proportional to 1/√n. To cut a mean's standard error in half, you need four times the sample size. Smaller SE means narrower confidence intervals and a lower probability of a Type II error.
Margin of error equals the critical value times the standard error, so for a one-sample t-interval it's t*(s/√n). SE is the raw measure of sampling variability; margin of error scales it up to match your confidence level.