Confidence Intervals — AP Statistics Stats Study Guide

For: AP Statistics candidates sitting AP Statistics.

Covers: Core confidence interval structure, calculation for one proportion, one mean, difference of two proportions/means, and common interpretation rules for AP exam success.

You should already know: Algebra 2, basic probability intuition.

A note on the practice questions: All worked questions in the "Practice Questions" section below are original problems written by us in the AP Statistics style for educational use. They are not reproductions of past College Board papers and may differ in wording, numerical values, or context. Use them to practise the technique; cross-check with official College Board mark schemes for grading conventions.

1. What Are Confidence Intervals?

A confidence interval (CI) is a range of plausible values for an unknown population parameter, calculated from sample data, that quantifies the uncertainty of your estimate. It is tied to a pre-specified confidence level (typically 90%, 95%, or 99%) that represents the long-run proportion of intervals that would capture the true population parameter if you repeated the sampling process infinitely. Common synonyms include interval estimate and confidence bounds. This topic is the foundation of inferential statistics, makes up 10-15% of the AP Statistics exam, and appears on at least one free response question (FRQ) in 75% of past exams per College Board data.

2. CI structure: estimate ± margin of error

All confidence intervals follow the same universal structure, no matter what parameter you are estimating: $$\text{Confidence Interval}=\text{Sample Statistic}\pm \text{Margin of Error}$$ We break down each component below:

1. Sample Statistic: Your point estimate of the population parameter, e.g., sample proportion $\hat{p}$, sample mean $\bar{x}$, or difference between two sample proportions ${\hat{p}}_1-{\hat{p}}_2$. This is your best single guess for the unknown parameter.
2. Margin of Error (ME): Accounts for random sampling variability (the natural variation in results across different samples of the same size). It is calculated as: $$\text{Margin of Error}=\text{Critical Value}\times \text{Standard Error}$$

• The critical value is tied to your confidence level: for a 95% confidence level using the normal distribution, the critical z-value $z^{\ast }=1.96$, meaning 95% of the sampling distribution lies within ±1.96 standard errors of the mean.
• The standard error (SE) is the standard deviation of the sampling distribution of your sample statistic, measuring how much the statistic is expected to vary across samples.

Worked Example

Suppose you calculate a sample proportion $\hat{p}=0.62$ from a random sample, with a standard error of 0.04, and use a 95% confidence level.

• Critical value $z^{\ast }=1.96$
• ME = $1.96\times 0.04=0.078$
• Final CI: $0.62\pm 0.078=(0.542,0.698)$

Note: The margin of error only accounts for random sampling error, not bias from nonresponse, measurement error, or unrepresentative sampling—this is a common exam multiple-choice question.

3. CI for one proportion

Use this interval to estimate the true proportion of a population with a specific characteristic (e.g., proportion of voters who support a candidate, proportion of students who pass a test).

First, check 3 required conditions (exam graders deduct 1-2 points for skipping these):

1. Random: The sample is randomly selected from the population, or groups are formed via random assignment in an experiment.
2. Independence: The sample size $n$ is less than 10% of the total population size (10% condition) to ensure individual observations are roughly independent when sampling without replacement.
3. Normal (Large Counts): The number of successes $n\hat{p}\ge 10$ and failures $n(1-\hat{p})\ge 10$, so the sampling distribution of $\hat{p}$ is approximately normal.

Formula

$$\hat{p}\pm z^{\ast }\sqrt{\frac{\hat{p}(1-\hat{p})}{n}}$$ Where $z^{\ast }$ is the critical z-value for your chosen confidence level.

Worked Example

A random sample of 200 high school students finds 112 report eating breakfast daily. Calculate a 95% confidence interval for the proportion of all high school students who eat breakfast daily.

1. Check conditions: Random (stated), 200 < 10% of all high school students, $112\ge 10$ and $200-112=88\ge 10$: all conditions met.
2. Calculate $\hat{p}=112/200=0.56$
3. SE = $\sqrt{\frac{0.56(1-0.56)}{200}}=\sqrt{0.001232}\approx 0.035$
4. ME = $1.96\times 0.035=0.069$
5. Final CI: $0.56\pm 0.069=(0.491,0.629)$

4. CI for one mean

Use this interval to estimate the true mean of a quantitative variable in a population (e.g., mean monthly rent for apartments, mean test score for a class).

Required conditions:

1. Random: Same as for proportions.
2. Independence: Same 10% condition applies.
3. Normal/Large Sample: Either the population is normally distributed, or the sample size $n\ge 30$ (Central Limit Theorem applies, so the sampling distribution of $\bar{x}$ is approximately normal). For $n<30$, confirm no strong skewness or outliers in the sample distribution.

Formula

In almost all real and exam scenarios, the population standard deviation $\sigma$ is unknown, so we use the sample standard deviation $s$ and the t-distribution (which has fatter tails than the normal distribution to account for extra uncertainty from estimating $\sigma$ with $s$): $$\bar{x}\pm t_{df=n-1}^{\ast }\frac{s}{\sqrt{n}}$$ Where $t^{\ast }$ is the critical t-value for your confidence level, with degrees of freedom $df=n-1$. You will be given a t-table on the AP exam to look up this value.

Worked Example

A random sample of 25 college students reports a mean monthly grocery spending of $320,withasamplestandarddeviationof$45. Calculate a 90% confidence interval for the mean monthly grocery spending of all college students.

1. Check conditions: Random (stated), 25 < 10% of college students, no strong skewness/outliers noted: all conditions met.
2. $df=25-1=24$, so $t^{\ast }$ for 90% confidence = 1.711
3. SE = $\frac{45}{\sqrt{25}}=9$
4. ME = $1.711\times 9=15.4$
5. Final CI: $320\pm 15.4=(\$304.6,\$335.4)$

5. CI for difference of two proportions / means

Use these intervals to estimate the difference between parameters for two independent populations (e.g., difference in pass rates between two classes, difference in mean salary between two industries).

5.1 CI for difference of two proportions ($p_1-p_2$)

Conditions:

Same as one-proportion CI, but apply to both samples, and confirm the two samples are independent of each other. The large counts condition requires $n_1{\hat{p}}_1,n_1(1-{\hat{p}}_1),n_2{\hat{p}}_2,n_2(1-{\hat{p}}_2)$ all ≥ 10.

Formula:

$$({\hat{p}}_1-{\hat{p}}_2)\pm z^{\ast }\sqrt{\frac{{\hat{p}}_1(1-{\hat{p}}_1)}{n_1}+\frac{{\hat{p}}_2(1-{\hat{p}}_2)}{n_2}}$$ If this interval does not include 0, you have evidence of a statistically significant difference between the two population proportions.

5.2 CI for difference of two means (${\mu }_1-{\mu }_2$)

Conditions:

Same as one-mean CI for both groups, and confirm samples are independent. Use the conservative degrees of freedom $df=\min (n_1-1,n_2-1)$ for manual calculation (your calculator will compute a more precise df automatically, but the conservative estimate is accepted on the AP exam).

Formula (unpooled, standard for AP Stats):

$$({\bar{x}}_1-{\bar{x}}_2)\pm t^{\ast }\sqrt{\frac{s_1^2}{n_1}+\frac{s_2^2}{n_2}}$$ Only use the pooled variance formula if the problem explicitly states the two populations have equal variances (this is very rare on the AP exam).

6. Interpretation cautions

Interpretation questions make up 30% of CI-related exam points, and graders are strict about wording.

Correct standard interpretation template:

"We are [confidence level]% confident that the true [population parameter] for [population of interest] is between [lower bound] and [upper bound]."

The confidence level itself is interpreted as:

"If we took 100 random samples of the same size from this population and calculated a [confidence level]% CI for each, approximately [confidence level] of those intervals would capture the true population parameter."

Common incorrect interpretations to avoid:

1. ❌ "There is a 95% chance the true parameter is in this interval": The parameter is a fixed number, not a random variable, so probability does not apply to it. Only the interval itself is random.
2. ❌ "95% of the population has values in this interval": The interval estimates the population parameter (mean/proportion), not individual observations.
3. ❌ "We are 95% confident the sample statistic is in this interval": The sample statistic is exactly at the center of the interval, so you know it is there with 100% certainty.
4. ❌ Generalizing to the wrong population: If your sample only includes California high schoolers, do not claim your interval applies to all US high schoolers.

7. Common Pitfalls (and how to avoid them)

• Pitfall 1: Skipping condition checks before calculating a CI. Why students do it: They want to jump straight to calculations. Correct move: Always list and verify all 3 conditions first, as College Board FRQ rubrics deduct 1-2 points for missing conditions.
• Pitfall 2: Using $z^{\ast }$ instead of $t^{\ast }$ for one-mean CIs when $\sigma$ is unknown. Why students do it: z-values are easier to memorize. Correct move: If you only have sample standard deviation $s$, use the t-distribution with $n-1$ df, even for large sample sizes.
• Pitfall 3: Confusing standard deviation and standard error. Why students do it: The terms look similar. Correct move: Standard deviation describes variability in individual data points, while standard error describes variability in the sample statistic—always use SE in CI formulas.
• Pitfall 4: Using pooled variance for two-sample mean CIs unnecessarily. Why students do it: Some older textbooks teach pooled variance as default. Correct move: Only use pooled variance if the problem explicitly states the two populations have equal variances.
• Pitfall 5: Forgetting that margin of error does not account for bias. Why students do it: They assume a small ME means the interval is accurate. Correct move: Always note that ME only accounts for random sampling error, not response bias, nonresponse bias, or sampling frame errors.

8. Practice Questions (AP Statistics Style)

Question 1

A local bakery randomly selects 40 days of sales records and finds a mean of 127 chocolate chip cookies sold per day, with a sample standard deviation of 22 cookies. (a) Calculate a 95% confidence interval for the mean number of chocolate chip cookies sold per day. (b) Interpret the interval in context.

Solution

(a) First check conditions: Random sample of days (stated), 40 < 10% of all operating days, $n=40\ge 30$ so CLT applies. $df=39$, $t^{\ast }=2.023$ for 95% confidence. SE = $\frac{22}{\sqrt{40}}\approx 3.48$. ME = $2.023\times 3.48\approx 7.04$. Final CI: $127\pm 7.04=(119.96,134.04)\approx (120,134)$. (b) We are 95% confident that the true mean number of chocolate chip cookies sold per day at the bakery is between 120 and 134.

Question 2

A survey of 300 randomly selected urban residents finds 189 support a new bike lane project, while a survey of 250 randomly selected suburban residents finds 125 support the project. Calculate a 90% confidence interval for the difference in support rates between urban and suburban residents (urban minus suburban).

Solution

Check conditions: Independent random samples (stated), both sample sizes <10% of their respective populations, $n_1{\hat{p}}_1=189,n_1(1-{\hat{p}}_1)=111,n_2{\hat{p}}_2=125,n_2(1-{\hat{p}}_2)=125$, all ≥10. ${\hat{p}}_u=0.63$, ${\hat{p}}_s=0.5$, difference = 0.13. SE = $\sqrt{\frac{0.63\ast 0.37}{300}+\frac{0.5\ast 0.5}{250}}\approx 0.042$. $z^{\ast }=1.645$ for 90% confidence. ME = $1.645\ast 0.042\approx 0.069$. Final CI: $0.13\pm 0.07=(0.06,0.20)$. We are 90% confident that urban support for the bike lane is 6 to 20 percentage points higher than suburban support.

Question 3

A student calculates a 95% confidence interval for the proportion of students at their school with a part-time job as (0.32, 0.48). They say "There is a 95% chance that the true proportion of students with part-time jobs is between 32% and 48%." Identify the error in their interpretation, and write a correct version.

Solution

Error: The student incorrectly assigns a probability to the fixed population parameter being in the interval. The true proportion is a constant, so it either is or is not in the calculated interval, no probability applies to it. Correct interpretation: We are 95% confident that the true proportion of students at the school with a part-time job is between 32% and 48%. This means if we took 100 random samples of the same size and calculated a 95% CI for each, about 95 of those intervals would capture the true proportion.

9. Quick Reference Cheatsheet

10. What's Next

Confidence intervals are directly linked to the next core AP Statistics topic: significance (hypothesis) testing. If a confidence interval for a parameter does not include a hypothesized value, you can reject that hypothesis at the corresponding significance level (e.g., a 95% CI corresponds to a 0.05 significance level test). Two-sample CIs also lay the foundation for chi-square tests for two-way tables and regression inference, which make up another 15-20% of the AP Stats exam. Mastering CI calculation and interpretation will cut your study time for these later topics in half, as the core logic of sampling variability and uncertainty applies across all inferential methods.

If you struggle with any of the concepts in this guide, from picking the correct critical value to interpreting intervals correctly, you can ask Ollie for personalized practice problems, step-by-step walkthroughs of past AP FRQs, or quizzes to test your mastery at any time. Head to Ollie, our AI tutor, to get immediate feedback tailored to your learning gaps, or browse more AP Statistics study guides on the homepage to build your full exam prep plan.