Null and Alternative Hypotheses in AP Statistics

In AP Statistics, the null hypothesis (H0) is a claim of no effect or no difference about a population parameter, and the alternative hypothesis (Ha) is the claim you're looking for evidence of. A significance test uses sample data to decide whether there's convincing evidence against H0 in favor of Ha.

What is Null and Alternative Hypotheses?

Every significance test starts with two competing claims about a population parameter (like a true proportion p or a true mean μ).

• The null hypothesis (H0) is the "nothing's going on" claim. It says there's no effect, no difference, or the parameter equals some specific value. Example: H0: p = 0.50.
• The alternative hypothesis (Ha) is the claim the study is trying to find evidence for. It says the parameter is greater than, less than, or not equal to that value. Example: Ha: p > 0.50.

Think of it like a courtroom. H0 is "innocent until proven guilty." You assume H0 is true, then ask how surprising your sample data would be if it were. If the data are surprising enough (a small p-value), you reject H0 and conclude there's convincing evidence for Ha. If not, you fail to reject H0. That doesn't mean H0 is proven true. It just means the evidence wasn't strong enough to rule it out.

Two rules make or break your hypotheses on the exam. First, they're always about parameters, never sample statistics (write p, not p̂). Second, the equals sign always lives in H0.

Why Null and Alternative Hypotheses matters in AP® Statistics

This term sits under Topic 3.6, p-Values, and it's the foundation that topic is built on. A p-value is defined as the probability of getting a result at least as extreme as yours assuming H0 is true. You can't compute it, or even interpret it, until you know what H0 says and which direction Ha points. The direction of Ha (>, <, or ≠) decides whether you look at one tail or both tails.

Beyond that one topic, hypotheses are the first step of every significance test in the course. That covers tests for proportions, means, differences, chi-square, and slopes. Getting them right (correct parameter, correct symbol, defined in context) is a scoring requirement on inference FRQs. So this isn't a one-topic vocabulary word. It's the opening move of the whole inference half of AP Stats. For the full walkthrough of how p-values work, head to the 3.6 p-Values study guide.

How Null and Alternative Hypotheses connects across the course

p-Value (Topic 3.6)

The p-value only makes sense in relation to H0. It answers one question: "If the null were true, how likely is data this extreme?" Ha tells you what "extreme" means, which is the right tail, the left tail, or both.

Significance Level (Inference units)

The significance level α is the cutoff you set before collecting data for how much evidence you need to reject H0. Compare the p-value to α. If p < α, reject H0. If not, fail to reject. The hypotheses say what's being decided, and α says how strict the judge is.

Type I and Type II Errors (Inference units)

Both errors are defined by the hypotheses. A Type I error rejects H0 when it's actually true, which is a false alarm. A Type II error fails to reject H0 when Ha is actually true, which is a missed effect. If you can't state H0 and Ha in context, you can't describe either error in context.

Standard Error (Inference units)

In a test, the standard error is calculated assuming H0 is true. That's why a one-proportion z-test plugs the null value p0 into the standard error formula instead of p̂. The null hypothesis literally shapes the math of the test statistic.

Is Null and Alternative Hypotheses on the AP® Statistics exam?

Expect hypotheses on nearly every inference question. Released FRQs that used this term include 2021 Q4 (a pet supply manager testing whether offering past customers $10 off encourages repeat purchases), 2023 Q4 (a study comparing an omega-3 supplement to a placebo for irritability), 2024 Q5 (Michelle sampling at a baseball card convention with about 20,000 attendees), and 2025 Q4.

Here's what you need to DO:

• State both hypotheses with correct symbols. Use a parameter (p, μ, p1 − p2, μdiff), put the equals sign in H0, and pick the right direction for Ha based on the research question.
• Define the parameter in context. "p = the true proportion of all past customers who make a repeat purchase" earns credit. A bare "p" usually doesn't.
• Connect your conclusion back to Ha. "Because the p-value of 0.012 < α = 0.05, reject H0. There is convincing evidence that..." in the language of Ha.

Multiple-choice stems often ask which pair of hypotheses fits a scenario. Common wrong answers use p̂ instead of p, or flip the direction of Ha.

Null and Alternative Hypotheses vs Null Hypothesis (H0) vs. Alternative Hypothesis (Ha)

The null is the default "no change" claim and always contains the equals sign (H0: μ = 12). The alternative is the claim the researcher suspects and uses >, <, or ≠ (Ha: μ < 12). A quick check is to ask what you're trying to find evidence FOR. That goes in Ha. You never try to prove H0. You only decide whether there's enough evidence against it.

Key things to remember about Null and Alternative Hypotheses

• The null hypothesis H0 states no effect or no difference, and it always contains the equals sign.

• The alternative hypothesis Ha is what you're looking for evidence of, and it uses >, <, or ≠ depending on the research question.

• Hypotheses are always written about population parameters like p or μ, never about sample statistics like p̂ or x̄.

• The p-value is calculated assuming H0 is true, so the null hypothesis sets the baseline for the entire test.

• You either reject H0 or fail to reject H0, and you never say you accept or prove the null.

• On FRQs, you need to define the parameter in context and state your conclusion in terms of Ha to earn full credit.

Frequently asked questions about Null and Alternative Hypotheses

What are null and alternative hypotheses in AP Stats?

They're the two competing claims in a significance test. H0 says there's no effect or the parameter equals a specific value (like p = 0.50), and Ha says the parameter is greater than, less than, or not equal to that value.

Can you accept the null hypothesis in AP Stats?

No. If the p-value is greater than α, you "fail to reject H0," which means the data didn't give convincing evidence against it. Saying you "accept" or "prove" H0 can cost you points on an FRQ.

How is the null hypothesis different from the alternative hypothesis?

The null is the default claim with an equals sign (H0: p = 0.30), and the alternative is the competing claim you want evidence for (Ha: p > 0.30). The test assumes H0 is true and checks whether the sample data make that assumption hard to believe.

Do you use p-hat or p when writing hypotheses?

Use p, the population parameter. Hypotheses are claims about the whole population, and you already know p̂ from your sample, so writing H0: p̂ = 0.50 is a classic mistake.

How do I know if Ha should be one-sided or two-sided?

Read the research question. If it asks whether something has increased or decreased, use > or <. If it asks whether something is simply different, use ≠, and the p-value will count both tails.