Skip to content

Unit 4 of 5

Inference for Quantitative Data: Means

15% of the exam206 reviewed questions45 key terms17 min read

Inference for Quantitative Data: Means is worth about 15% of the AP Statistics exam. Our bank tests it with 206 reviewed questions across 45 distinct ideas — 50 easier, 126 medium and 30 harder.

On this page · 12 sections

This unit is about estimating and testing a population mean when all you have is one sample. You want to know the true mean commute time, battery life or score gain, which is μ. You only see the sample mean, x̄, and x̄ changes from sample to sample. Every tool in the unit rests on one question: how much would x̄ bounce around if you took another random sample? The answer is the sampling distribution of x̄, and its spread is the standard error.

Once you know how far x̄ typically lands from μ, two procedures follow. A confidence interval runs that logic forward: start at x̄ and reach out a few standard errors to capture plausible values of μ. A significance test runs it backward: assume a value of μ, then ask how surprising your x̄ would be if that value were true. Both use the t distribution instead of the normal curve, because you almost never know the population standard deviation σ and must plug in the sample standard deviation s.

The same machinery handles three designs. A one-sample t procedure studies one mean. A paired t procedure studies the mean of differences when each value has a natural partner. A two-sample t procedure compares the means of two independent groups. The exam rewards you for picking the right design, checking conditions, doing the calculation, and then writing a conclusion in context that says no more than the data allow.

The unit in a minute

One line per topic. Tap any line for the full notes.

  1. The sampling distribution of x̄ and the Central Limit Theoremx̄ is centred at μ with spread σ/√n, and it is normal if the population is normal or approximately normal when n is large.
  2. Standard error, the t distribution and degrees of freedomUse s/√n as the standard error and the t distribution with df = n − 1, because you are estimating σ with s.
  3. Conditions: random, independent, normal enoughState and check random, 10%, and normal/large-sample conditions in context, and apply normality to the differences when data are paired.
  4. Building and interpreting a one-sample t intervalInterpret the interval as plausible values for the population mean in context, and explain the confidence level as the long-run capture rate of the method.
  5. The one-sample t test, from hypotheses to conclusionState hypotheses about μ, compute t = (x̄ − μ₀)/(s/√n), find the p-value in the direction of Hₐ, and conclude in context in terms of evidence.
  6. Paired data versus two independent samplesIf each value has a partner, analyse the differences as one sample; if the groups are separate, use the two-sample standard error √(s₁²/n₁ + s₂²/n₂).
  7. Duality, statistical significance and practical significanceAn interval that excludes the null value matches a two-sided rejection, and the interval's size, not the p-value, tells you whether the effect matters.
Grind Inference for Quantitative Data: Means questionsAlready know this? Skip the notes and prove it.

The big ideas

x̄ is a random variable with a predictable shape

Before you take the sample, x̄ is unknown, and its possible values form a distribution centred at μ, which is why x̄ is called unbiased. Its standard deviation is σ/√n, so bigger samples give less variable means. Its shape is normal if the population is normal, and approximately normal for large n by the Central Limit Theorem. Every interval and test in this unit borrows these three facts.

Inference measures distance in standard errors

A test statistic and a margin of error both ask the same thing: how many standard errors separate two numbers? In a test, t = (x̄ − μ₀)/(s/√n) counts the standard errors between your sample mean and the claimed mean. In an interval, the margin of error is t* standard errors wide. If you understand standard error, the formulas stop being things to memorise.

The design decides the procedure

Paired data and two independent groups can look alike in a table, but they need different math. If the same subjects are measured twice, or subjects are matched in pairs, you subtract within each pair and analyse one list of differences. If two separate random samples or two randomly assigned groups are involved, you use the two-sample formula. Picking the wrong one changes the standard error and the degrees of freedom, so it can flip a conclusion.

Evidence is not proof, and significance is not importance

A small p-value says your data would be rare if H₀ were true. It does not give the probability that H₀ is true, and it does not say the effect is large. A large p-value means you lack evidence, not that H₀ has been proven. Whether a real effect is big enough to matter is a separate question about effect size and context, so a 2-point gain can be statistically significant and still useless.

Study notes

7 topics. Each one opens with its key idea, then the detail you need to use it.

1. The sampling distribution of x̄ and the Central Limit Theorem

Key idea: x̄ is centred at μ with spread σ/√n, and it is normal if the population is normal or approximately normal when n is large.

Picture taking every possible random sample of size n from a population and recording each x̄. Those values form the sampling distribution of x̄. Its mean equals the population mean μ, so x̄ is an unbiased estimator, and its standard deviation is σ/√n, which shrinks as n grows. Because of the square root, you must quadruple the sample size to cut the spread in half.

Shape depends on the population. If the population is normal, x̄ is exactly normal for any n, even n = 4. If the population is skewed or unknown, the Central Limit Theorem says x̄ becomes approximately normal as n grows, and n ≥ 30 is the usual rule of thumb. The CLT is about the distribution of sample means. It does not make the population or a single sample look normal.

This lets you compute probabilities about x̄. Suppose a population has μ = 40 and σ = 12, and you take n = 36. The standard deviation of x̄ is 12/√36 = 2. To find P(x̄ > 43), compute z = (43 − 40)/2 = 1.5, so the probability is about 0.067. Notice that you divide by σ/√n, not by σ. Using σ = 12 would answer a question about one individual, not about a mean of 36 people.

2. Standard error, the t distribution and degrees of freedom

Key idea: Use s/√n as the standard error and the t distribution with df = n − 1, because you are estimating σ with s.

The formula σ/√n needs σ, the population standard deviation, which you almost never know. So you estimate it with s, the sample standard deviation, and the result, s/√n, is called the standard error of x̄, which estimates how far x̄ typically falls from μ.

Swapping σ for s adds extra uncertainty, because s is itself random. Small samples can give an s that is too small, which would make you overconfident. The t distribution corrects for this. It is centred at 0 and bell-shaped like the standard normal, but its heavier tails mean critical values are larger, so intervals are wider and p-values are bigger than z would give.

Which t curve you use depends on the degrees of freedom. For one sample or for paired differences, df = n − 1. As df increases, s becomes a more reliable estimate of σ and the t curve moves closer to the normal curve. For df = 24, the 95% critical value is t* = 2.064, while z* is 1.96. For df = 5, t* jumps to 2.571.

This is why the exam cares about the label. If a stem gives s and asks for the test statistic, the answer is a t statistic. Calling it z is wrong even when the arithmetic matches, because z assumes σ is known.

3. Conditions: random, independent, normal enough

Key idea: State and check random, 10%, and normal/large-sample conditions in context, and apply normality to the differences when data are paired.

Every t procedure rests on three conditions, and each one protects a different part of the method. Check them before you calculate, and tie each check to the data in the stem.

Random. The data must come from a random sample or a randomized experiment. Without randomness, the sampling distribution does not exist in any useful sense, and no formula fixes the bias. A convenience sample from one university can give a significant result that still does not generalize to all students.

Independent. When sampling without replacement, the sample should be no more than 10% of the population. This keeps the observations close to independent, which the formula s/√n assumes. In an experiment, random assignment gives independence between groups.

Normal or large sample. The sampling distribution of x̄ must be approximately normal. You can claim this three ways. The stem says the population is normal. The sample size is at least 30, so the CLT applies. Or, with a small sample, a dotplot, boxplot or normal probability plot of the data shows no strong skew and no outliers. With n = 12 and a big outlier, a t procedure is not trustworthy, because the mean and s are both pulled by that one value.

For paired data, check normality on the differences, not on the two original lists. For two samples, check each group separately.

4. Building and interpreting a one-sample t interval

Key idea: Interpret the interval as plausible values for the population mean in context, and explain the confidence level as the long-run capture rate of the method.

A confidence interval has the form statistic ± margin of error. For a mean, that is x̄ ± t*·(s/√n). Here is a full calculation. A random sample of 25 commuters has x̄ = 31.4 minutes and s = 6.5 minutes. The standard error is 6.5/√25 = 1.3. With df = 24, the 95% critical value is 2.064, so the margin of error is 2.064 × 1.3 ≈ 2.68. The interval is about 28.7 to 34.1 minutes.

A good interpretation names the parameter in context. You would write that you are 95% confident the true mean commute time for all commuters in this city is between 28.7 and 34.1 minutes. Name the population mean, not the sample mean. x̄ is already known (31.4) because you computed it from the data, and the interval is built around it, so there is nothing to estimate about it.

The confidence level describes the method, not this one interval. If you took many random samples and built an interval from each, about 95% of those intervals would capture μ. This specific interval either contains μ or it does not, so you should not say there is a 95% probability that μ is in it.

Three things control the width. A higher confidence level needs a bigger t*, so the interval widens. A larger sample shrinks s/√n, so the interval narrows. More variable data means a larger s and a wider interval. The margin of error only covers random sampling error. It does nothing about bias from a bad sampling method or from nonresponse.

5. The one-sample t test, from hypotheses to conclusion

Key idea: State hypotheses about μ, compute t = (x̄ − μ₀)/(s/√n), find the p-value in the direction of Hₐ, and conclude in context in terms of evidence.

A test starts with two hypotheses about the parameter. H₀ gives a specific value, such as μ = 10 hours. Hₐ says μ is less than, greater than, or not equal to that value. The direction comes from the research question, and you must choose it before you look at the data.

Say a company claims its batteries last 10 hours on average, and a reviewer suspects they last less. A random sample of 16 batteries gives x̄ = 9.4 hours and s = 1.2 hours. The hypotheses are H₀: μ = 10 and Hₐ: μ < 10. The standard error is 1.2/√16 = 0.3, so the test statistic is t = (9.4 − 10)/0.3 = −2.0 with df = 15.

The p-value is the probability of getting a sample mean at least this far below 10, assuming the true mean really is 10. Because Hₐ is one-sided, you only use the lower tail, and p ≈ 0.032. A two-sided Hₐ would double it to about 0.064.

Compare p to α: here 0.032 < 0.05, so you reject H₀. The conclusion should say there is convincing evidence that the true mean battery life of these batteries is less than 10 hours. If p had been above α, you would fail to reject H₀ and say there is not convincing evidence. You never "accept" H₀, and rejecting it does not prove it false. A test weighs evidence, so the wording must stay at the level of evidence.

6. Paired data versus two independent samples

Key idea: If each value has a partner, analyse the differences as one sample; if the groups are separate, use the two-sample standard error √(s₁²/n₁ + s₂²/n₂).

Paired data arise when each observation has a natural partner. Examples include the same student before and after tutoring, the left and right eye of one patient, or twins split between two treatments. Pairing removes the variation between individuals, so the test is often more sensitive.

To analyse paired data, subtract within each pair to get one list of differences. Then run a one-sample t procedure on those differences, with df = number of pairs − 1. Suppose 12 students take a quiz before and after a review session, and the mean gain is 3.5 points with s_d = 4.2. The standard error is 4.2/√12 ≈ 1.21, so t ≈ 2.89 with df = 11. The one-sided p-value is about 0.007. Fifteen pairs give df = 14, not 28, because you have 15 differences, not 30 separate values.

Two independent samples arise when the groups contain different individuals, picked or assigned separately. The standard error combines both groups: SE = √(s₁²/n₁ + s₂²/n₂). The statistic is t = (x̄₁ − x̄₂)/SE when H₀ says the means are equal. Your calculator finds df with a complex formula, but a safe hand method uses the smaller of n₁ − 1 and n₂ − 1, which gives a slightly larger p-value.

For example, a researcher takes random samples of students from two large schools and compares their mean scores on the same test. School 1 has n = 20, x̄ = 78, s = 10, and school 2 has n = 25, x̄ = 72, s = 12. The hypotheses are H₀: μ₁ = μ₂ and Hₐ: μ₁ ≠ μ₂, where μ₁ and μ₂ are the true mean scores at each school. The SE is √(5 + 5.76) ≈ 3.28, so t ≈ 1.83. With conservative df = 19, the two-sided p-value is about 0.08. Because 0.08 > 0.05, there is not convincing evidence that the true mean scores at the two schools differ.

To estimate the difference instead of testing it, use the two-sample t interval (x̄₁ − x̄₂) ± t*·√(s₁²/n₁ + s₂²/n₂). Its interpretation names the difference in population means, μ₁ − μ₂, in context, such as the true mean score at school 1 minus the true mean score at school 2.

7. Duality, statistical significance and practical significance

Key idea: An interval that excludes the null value matches a two-sided rejection, and the interval's size, not the p-value, tells you whether the effect matters.

A two-sided test at level α and a confidence interval at level 1 − α give matching answers. If a 95% interval for a mean difference is 5 to 11 points, then 0 is not a plausible value. So a two-sided test of H₀: μ_d = 0 would reject at α = 0.05. If the interval had been −1 to 6, 0 would be plausible and the test would fail to reject. You do not need the p-value to make this call, but the interval must match the test: 95% with α = 0.05, 99% with α = 0.01.

The interval also tells you something the p-value cannot: how big the effect is. That is where practical significance comes in. Statistical significance only says that a difference this large would rarely happen by chance if the true effect were zero, so the data give convincing evidence that the effect is not zero. With a huge sample, even a tiny difference gives a small p-value, because the standard error shrinks with √n. A drug that lowers cholesterol by 2 mg/dL might reach p = 0.03 yet be medically pointless when levels run 100 to 200 mg/dL.

Effect size is the size of the difference in real units, or relative to the spread of the data. Judge it against the context, not against α. An interval like 0.5 to 3.5 points on a 100-point test is significant but small. An interval like 8 to 20 points is both significant and large.

The last check is scope. A significant two-sample result from a convenience sample still does not generalize to a wider population. Only random assignment supports a cause-and-effect claim.

Common mistakes

Drawn from the wrong answers students actually pick on these questions.

Reading the p-value as the probability that H₀ is true, or turning p = 0.002 into "a 99.8% chance the treatment works."

The p-value assumes H₀ is true and then asks how often chance alone would give a result this extreme or more. It is a probability about the data under H₀, not about the hypothesis. It also says nothing about how often individuals improve.

Saying there is a 95% probability that the true mean lies in this interval, or that x̄ lies in it with 95% probability.

The true mean is a fixed number, and x̄ is the centre of the interval, so neither is random once you have the data. The 95% describes the method: about 95% of intervals built this way capture μ. Say "we are 95% confident" and name the population mean in context.

Writing "accept H₀" when p > α, or "H₀ is definitely false" when p < α.

A large p-value means the evidence is not strong enough, so you fail to reject H₀. It does not prove H₀. A small p-value gives convincing evidence against H₀, which is not certainty. A p-value above α also does not make the test invalid; it only means you cannot reject.

Building the test statistic wrong: dividing by s instead of s/√n, dividing by n instead of √n, or flipping the subtraction to μ₀ − x̄.

Use t = (x̄ − μ₀)/(s/√n). The denominator is the standard error, which shrinks with the square root of the sample size. With x̄ = 5.8, μ₀ = 5, s = 1.8 and n = 36, the SE is 1.8/6 = 0.3, not 1.8 or 0.05.

Labelling the statistic z when the stem gives the sample standard deviation s.

When σ is unknown and you plug in s, the statistic follows a t distribution with n − 1 degrees of freedom. The arithmetic can match a z calculation, but the name and the p-value come from t.

Calling a significant result important, or deciding significance from the width of an interval or the sign of x̄.

Significance asks whether the null value is plausible. For a 95% interval, check whether it contains the null value, such as 0. Width shows precision, and a positive x̄ alone proves nothing. Whether the effect matters is a separate judgment about its size in context, which a tiny p-value cannot settle.

How this unit shows up on the exam

Multiple-choice questions in this unit come in a few recognisable shapes. Some give μ, σ and n and ask for the mean and standard deviation of x̄, or for a probability like P(x̄ > 24). Those reward dividing by √n and knowing when the shape is normal. Others give x̄, s and n and ask for the test statistic. The wrong choices are built from forgetting the square root, flipping the sign, or calling it z. A large group gives a p-value or an interval and offers several interpretations. Only one keeps the probability on the method or the data under H₀, and the rest assign probability to the hypothesis or the parameter. Expect questions that hand you a p-value and α and ask for the conclusion, with distractors that say "accept" or "proves". Design questions describe a study and ask whether it is paired or two-sample, or ask for the degrees of freedom. Watch for duality stems that give only an interval and ask about significance, and for scenarios that separate statistical from practical significance.

On free-response, means inference usually appears as a full procedure. Readers look for four parts. First, name the procedure and define the parameter in context, such as μ_d = the true mean difference in quiz score (after − before). Second, check the conditions with reference to the data. Third, show the mechanics: the test statistic or interval, the df, and the p-value. Fourth, write a conclusion that links the p-value to α and answers the question in context with evidence language. A correct number with a vague conclusion loses credit. So does a conclusion about x̄ instead of μ, or a paired study analysed as two independent groups. Interpretation questions about a confidence level or a p-value also show up as short parts, and they are graded on precise wording.

How the exam actually asks it

Real items from the AP Statistics bank, easier first. The answer is marked, and every wrong choice is explained — because knowing why a distractor is tempting is most of the work.

Worked example 1Easier
A one-sample t-test finds t = 2.31 with df = 24 and a two-sided p-value = 0.030. Using α = 0.05, what should the researcher conclude?
  1. AReject H0; the null hypothesis is definitively false.
  2. BFail to reject H0; there is not enough evidence against it.
  3. CReject H0; there is significant evidence against H0.Correct
  4. DAccept H0; the null hypothesis is proven true.

Why C is right

Since p = 0.030 < α = 0.05, we reject H0. There is statistically significant evidence at the 5% level that the true mean differs from the null value. Note that 'rejecting' H0 does not mean it is definitively false — it means the data provide sufficient evidence against it at this significance level.
Why the other 3 choices miss
A
'Definitively false' overstates the conclusion. Statistical testing provides evidence, not certainty. The correct wording is that there is significant evidence against H0.
B
Failing to reject requires p > α. Since p = 0.030 < α = 0.05, the evidence IS sufficient to reject H0 at the 5% significance level.
D
We never 'accept' or 'prove' H0. Failing to reject H0 means insufficient evidence to reject it. Here, we reject H0.
Worked example 2Easier
A random variable X has mean μ = 50 and standard deviation σ = 10. A simple random sample of n = 100 observations is taken. What are the mean and standard deviation of the sampling distribution of x-bar?
  1. AMean = 50, SD = 10.
  2. BMean = 5, SD = 1.
  3. CMean = 50, SD = 1.Correct
  4. DMean = 50, SD = 0.1.

Why C is right

The sampling distribution of x-bar has mean μ_x-bar = μ = 50 and standard deviation (standard error) σ_x-bar = σ/sqrt(n) = 10/sqrt(100) = 10/10 = 1. The probability follows from counting outcomes, conditional probability, or the stated random process.
Why the other 3 choices miss
A
Incorrect because the standard deviation of the sampling distribution of x-bar is σ/sqrt(n) = 1, not the population standard deviation σ = 10. The sampling distribution has less variability than the population.
B
Incorrect because the mean of the sampling distribution equals the population mean μ = 50, not 5. 5 might result from dividing μ by n (50/100 = 0.5) or computing something else incorrectly.
D
Incorrect because σ/sqrt(n) = 10/sqrt(100) = 1, not 0.1. 0.1 would result from σ/n = 10/100 = 0.1 (dividing by n instead of sqrt(n)).
Worked example 3Easier
The sampling distribution of x-bar for samples of n = 16 drawn from a normal population N(50, 8) has what mean and standard deviation?
  1. AMean = 50, SD = 2.Correct
  2. BMean = 50, SD = 8.
  3. CMean = 3.125, SD = 0.5.
  4. DMean = 50, SD = 4.

Why A is right

The population is N(50, 8) — mean 50, standard deviation 8 (not variance). Standard error of x-bar = σ/sqrt(n) = 8/sqrt(16) = 8/4 = 2. Sampling distribution: N(50, 2). The probability follows from counting outcomes, conditional probability, or the stated random process.
Why the other 3 choices miss
B
Incorrect because SD = 8 is the population standard deviation, not the standard error of x-bar. The SE = σ/sqrt(n) = 8/4 = 2.
C
Incorrect because 3.125 = 50/16 and 0.5 = 8/16 — these use division by n instead of sqrt(n). The mean of x-bar = μ = 50 (not μ/n) and SE = σ/sqrt(n) = 2.
D
Incorrect because SD = 4 = σ/sqrt(n) only if n = 4 (8/sqrt(4) = 4). With n = 16, SE = 8/sqrt(16) = 8/4 = 2.
Worked example 4Easier
When performing a two-sample t-test, a researcher chooses to use Welch's t-test rather than the pooled t-test. What assumption is relaxed by this choice?
  1. ARandom sampling from each population
  2. BEqual population variancesCorrect
  3. CIndependence of the two samples
  4. DNormality of the population distributions

Why B is right

Welch's t-test does not assume equal population variances (σ₁² = σ₂²). The pooled t-test requires this assumption and uses a combined variance estimate. Welch's method estimates the variance separately for each group and adjusts the degrees of freedom accordingly, making it more robust when variances differ.
Why the other 3 choices miss
A
Incorrect because random sampling is a fundamental design requirement for valid inference in both tests; Welch's method does not change this requirement.
C
Incorrect because independence between the two samples is required for both Welch's and the pooled two-sample t-test; neither relaxes this assumption.
D
Incorrect because normality of population distributions (or large n by CLT) is a condition for both forms of two-sample t-tests, not something Welch's test removes.

Can you do all of these?

Tick what you can do without looking. Anything left unticked is where to go back to.

0 of 10 ticked

Ready to test yourself?

Open practice pulls from the 36 free questions in this unit. The other 170 come with the one-time AP purchase, which unlocks every AP course.

Official topic list

The official course framework breaks this unit into 10 numbered topics.

  1. 4.1Sampling Distributions for Sample Means
  2. 4.2Constructing a Confidence Interval for a Population Mean or Population Mean Difference
  3. 4.3Justifying a Claim Based on a Confidence Interval for a Population Mean or Population Mean Difference
  4. 4.4Setting Up a Test for a Population Mean or Population Mean Difference
  5. 4.5Carrying Out a Test for a Population Mean or Population Mean Difference
  6. 4.6Sampling Distributions for the Difference Between Two Sample Means
  7. 4.7Constructing a Confidence Interval for the Difference Between Two Population Means
  8. 4.8Justifying a Claim Based on a Confidence Interval for the Difference Between Two Population Means
  9. 4.9Setting Up a Test for the Difference Between Two Population Means
  10. 4.10Carrying Out a Test for the Difference Between Two Population Means

Key terms in Inference for Quantitative Data: Means

Ordered by how often our reviewed questions test them. Open any term to see the questions that use it.

The rest of AP Statistics

Every unit with its share of the exam. You are on Unit 4.

  1. 1Exploring One-Variable Data and Collecting Data25% of exam
  2. 2Probability, Random Variables, and Probability Distributions20% of exam
  3. 3Inference for Categorical Data: Proportions20% of exam
  4. 4Inference for Quantitative Data: Means15% of exam
  5. 5Regression Analysis15% of exam