The alpha level (α), or significance level, is the threshold you set before a hypothesis test for deciding whether to reject the null hypothesis. If the p-value is less than α, you reject H₀. Alpha also equals the probability of a Type I error, so α = 0.05 means a 5% chance of rejecting a true null hypothesis.
The alpha level, written α and also called the significance level, is the cutoff you choose before running a hypothesis test. It answers one question: how small does my p-value have to be before I'm willing to reject the null hypothesis? If p ≤ α, you reject H₀. If p > α, you fail to reject H₀. That's the entire decision rule, and it shows up in nearly every inference problem in AP Stats.
Here's the part that makes α more than just a cutoff. Alpha is also the probability of a Type I error, meaning the chance you reject a null hypothesis that was actually true. Setting α = 0.05 means you've accepted a 5% risk of crying wolf, of declaring an effect exists when it doesn't. The default on the AP exam is 0.05, but problems sometimes use 0.01 or 0.10, and the choice is a trade-off. A smaller α makes Type I errors rarer but makes it harder to detect real effects (more Type II errors, lower power). A bigger α does the reverse.
Alpha lives in the inference half of AP Stats, starting with significance tests for proportions in Unit 6 and carrying through tests for means (Unit 7), chi-square tests (Unit 8), and slope tests (Unit 9). The CED expects you to interpret the p-value, compare it to a stated significance level, and justify a conclusion in context. You also need to connect α to errors, because the CED treats Type I error probability and α as the same number. Alpha even links back to confidence intervals, since a two-sided test at α = 0.05 corresponds to a 95% confidence interval. If you can explain what α = 0.05 actually means, you can handle the error-and-power questions that trip up most test-takers' classmates.
Type I Error (Units 6-7)
Alpha IS the probability of a Type I error. They're the same number wearing two hats. When you set α = 0.05, you're literally choosing to tolerate a 5% chance of rejecting a true null hypothesis. Exam questions love asking you to interpret this in context.
P-value (Units 6-9)
Alpha and the p-value are the two halves of every test decision. Alpha is the bar you set before collecting data; the p-value is what your data actually produced. The comparison p ≤ α versus p > α drives the conclusion in every inference FRQ.
Confidence Level (Units 6-7)
Confidence level and alpha are complements. A 95% confidence interval pairs with α = 0.05 for a two-sided test, because confidence level = 1 − α. If a 95% interval for a difference misses zero, the matching two-sided test at α = 0.05 rejects H₀.
Hypothesis Test (Units 6-9)
Alpha is one of the standard pieces of the four-step hypothesis test template. You state it up front with your hypotheses, then use it in the conclusion step. The same α logic repeats whether you're testing proportions, means, chi-square counts, or a regression slope.
On multiple choice, α shows up in questions asking which conclusion follows from a given p-value, what the probability of a Type I error is, or what happens to power and error rates when α changes. On FRQs, almost every significance-test problem states a significance level (usually α = 0.05) and grades your conclusion on whether you explicitly compared the p-value to α and answered in context. The rubric phrasing you want is something like "Because the p-value of 0.012 is less than α = 0.05, we reject H₀ and have convincing evidence that..." Skipping the comparison to α is one of the most common ways to lose a point on an otherwise correct test. Also be ready for error-interpretation parts, where you describe a Type I error in context and identify its probability as α.
Alpha is the threshold you pick before the test; the p-value is the result your data gives you after the test. Alpha is fixed (like 0.05) and represents the Type I error rate you'll tolerate. The p-value varies from sample to sample and measures how surprising your data would be if H₀ were true. You compare them to decide, but they answer different questions. Saying "the p-value is the probability the null is true" or treating α as something the data produces are both classic AP Stats mistakes.
Alpha (α) is the significance level, the cutoff chosen before a hypothesis test, and the AP default is α = 0.05.
The decision rule is simple. If the p-value is less than or equal to α, reject H₀; if it's greater than α, fail to reject H₀.
Alpha equals the probability of a Type I error, so α = 0.05 means a 5% chance of rejecting a null hypothesis that is actually true.
Lowering α (say to 0.01) makes Type I errors less likely but increases the chance of a Type II error and reduces the power of the test.
Confidence level and alpha are linked by confidence level = 1 − α, so a 95% confidence interval matches a two-sided test at α = 0.05.
On FRQs, always state the comparison explicitly, such as "since p = 0.03 < α = 0.05, we reject H₀," then give your conclusion in context.
Alpha (α) is the significance level of a hypothesis test, the threshold you compare the p-value to. If the p-value is at or below α, you reject the null hypothesis. It also equals the probability of a Type I error, and the most common value on the AP exam is 0.05.
No. A p-value below α gives you convincing evidence against H₀, but it never proves anything. With α = 0.05, about 5% of tests on a true null will reject it anyway just by random chance. That's why AP rubrics want "convincing evidence," not "proof."
Alpha is the fixed cutoff you choose before the test (like 0.05), and it's the Type I error rate you're willing to accept. The p-value comes from your sample data and measures how likely results at least as extreme would be if H₀ were true. You compare the p-value to α to make the decision.
It's a convention, not a law. A 5% Type I error rate balances catching real effects against falsely rejecting true nulls. AP problems sometimes use 0.01 when a false rejection would be costly, or 0.10 when missing a real effect would be worse. If a problem states a different α, use that one.
You make Type I errors less likely (only a 1% chance of rejecting a true null), but the trade-off is a higher chance of a Type II error and lower power. In plain terms, you're less likely to cry wolf but more likely to miss a real effect. This trade-off is a favorite multiple-choice topic.
Connect this key term to the AP exam workflow: review the course, practice questions, and check related study tools.
Review units, study guides, and course resources.
Check this vocabulary in multiple-choice context.
Apply key concepts in written AP responses.
Estimate the exam score you are working toward.
Review the highest-yield facts before practice.
Put the full course together before test day.