---
title: "Significance Level (Alpha) — AP Stats Definition & Guide"
description: "Alpha (α) is the cutoff probability, usually 0.05, you compare a p-value to when deciding to reject H₀. It anchors every hypothesis test in AP Stats Units 6-9."
canonical: "https://fiveable.me/ap-stats/key-terms/significance-level-alpha"
type: "key-term"
subject: "AP Statistics"
unit: "Unit 3"
---

# Significance Level (Alpha) — AP Stats Definition & Guide

## Definition

The significance level (alpha, α) is the threshold probability, typically 0.05, set before a hypothesis test; if the p-value is less than α, you reject the null hypothesis. Alpha also equals the probability of making a Type I error (rejecting a true null hypothesis).

## What It Is

The [significance level](/ap-stats/key-terms/significance-level "fv-autolink"), written as alpha (α), is the cutoff you choose **before** running a hypothesis test. It answers one question. How unlikely does my [sample](/ap-stats/unit-1/random-sampling-data-collection/study-guide/nQz8XwRMmIKKBS59qrew "fv-autolink") result need to be (assuming the null hypothesis is true) before I'm willing to reject that null hypothesis? The standard choice is α = 0.05, though problems sometimes use 0.01 or 0.10.

Here's the intuitive version. Alpha is the line in the sand. Once you collect data and compute a p-value, you compare the two numbers. If the p-value is less than α, your result is "statistically significant" and you reject H₀. If the p-value is greater than or equal to α, you fail to reject H₀. Alpha is also your accepted risk of a false alarm. If you set α = 0.05, you're agreeing to a 5% [chance](/ap-stats/unit-3 "fv-autolink") of rejecting a null hypothesis that was actually true.

## Why It Matters

[Alpha](/ap-stats/key-terms/alpha "fv-autolink") first shows up in Topic 6.4, Setting Up a Test for a Population Proportion, alongside learning objective [AP Stats](/ap-stats "fv-autolink") 6.4.A (identifying the null and alternative hypotheses). The CED treats the null hypothesis as the situation assumed correct unless evidence suggests otherwise, and alpha is what defines "enough evidence." Without a stated significance level, a p-value has nothing to be compared against, so your test has no decision rule.

It matters far beyond Unit 6. Every inference procedure on the exam (tests for [proportions](/ap-stats/unit-1/representing-categorical-variable-with-tables/study-guide/JUZVd7cRAnbarZyNoEAg "fv-autolink"), means, chi-square tests, and slope) uses the same p-value versus alpha comparison. If you can write the comparison sentence correctly once, you can write it for every test in the course. For the full setup of a proportion test, head to the Topic 6.4 study guide.

## Connections

### [Null Hypothesis (Unit 6)](/ap-stats/key-terms/null-hypothesis)

Alpha only makes sense relative to H₀. The null is the assumed-true baseline, and alpha measures how surprising your data must be, under that baseline, before you abandon it. No null, no alpha.

### [1-Prop Z-Test (Unit 6)](/ap-stats/key-terms/1-prop-z-test)

This is where alpha gets used in practice. The one-sample [z-test](/ap-stats/key-terms/z-test "fv-autolink") for a proportion produces a p-value, and your final decision is literally one comparison. Is the p-value less than α? That single inequality drives the entire conclusion sentence.

### [Type I Error (Unit 6)](/ap-stats/key-terms/type-i-error)

Alpha isn't just a cutoff, it IS the probability of a [Type I error](/ap-stats/key-terms/type-i-error "fv-autolink"). Setting α = 0.05 means accepting a 5% chance of rejecting a true null hypothesis. Lowering alpha makes false alarms rarer but makes it harder to detect real effects, which is the alpha-power tradeoff tested later in Unit 6.

### Confidence Intervals (Units 6-9)

Alpha and confidence level are two sides of the same coin. A 95% confidence interval corresponds to α = 0.05 for a two-sided test. A two-sided test rejects H₀ exactly when the hypothesized value falls outside the matching confidence interval.

## On the AP Exam

Multiple-choice questions test whether you know what alpha means, not just how to use it. A classic stem gives you a p-value and a significance level and asks for the correct conclusion, or asks what a p-value represents (the probability of getting a result at least as extreme as the observed one, assuming H₀ is true) so you can tell it apart from alpha. On FRQs, nearly every inference question requires the comparison explicitly. Graders look for a sentence like "Because the p-value of 0.032 is less than α = 0.05, we reject H₀ and have convincing evidence that..." written in context. Two habits earn points. State alpha before computing anything, and never write "accept H₀." The correct phrasing when p ≥ α is "fail to reject H₀."

## Significance level (alpha) vs P-value

Alpha is chosen before the test; the p-value is computed from your data after. Alpha is the fixed standard of evidence (usually 0.05), while the p-value measures how surprising your actual sample is if H₀ were true. You compare them to decide. A small p-value (less than α) means reject H₀. Mixing these up, like saying "the p-value is the chance of a Type I error," is one of the most common ways to lose FRQ points.

## Key Takeaways

- The significance level α is the cutoff probability, set before the test, that determines whether a p-value counts as convincing evidence against the null hypothesis.
- The decision rule is simple. If the p-value is less than α, reject H₀; if the p-value is greater than or equal to α, fail to reject H₀.
- Alpha equals the probability of a Type I error, meaning rejecting a null hypothesis that is actually true.
- The default is α = 0.05, but if an FRQ doesn't state a significance level, you should explicitly say you're using 0.05.
- Never say "accept the null hypothesis" on the exam; failing to reject H₀ just means the evidence wasn't strong enough.
- A confidence level and alpha add to 100%, so a 95% confidence interval pairs with a two-sided test at α = 0.05.

## FAQs

### What is the significance level (alpha) in AP Stats?

It's the threshold probability, usually 0.05, that you set before a hypothesis test. If your p-value comes out below alpha, you reject the null hypothesis; otherwise you fail to reject it.

### Is alpha the same thing as the p-value?

No. Alpha is a fixed cutoff you choose before collecting data, while the p-value is calculated from your sample. The test decision comes from comparing the two, and confusing them is a frequent point-loser on FRQs.

### Does a result above alpha prove the null hypothesis is true?

No. When the p-value is at or above α, you fail to reject H₀, which only means the evidence wasn't strong enough. It never proves H₀ is correct, which is why "accept H₀" is wrong phrasing on the exam.

### Why is alpha usually 0.05?

It's a convention, not a law. α = 0.05 means tolerating a 5% chance of a Type I error, which balances catching real effects against false alarms. Problems use 0.01 when false alarms are costly and 0.10 when missing a real effect is worse.

### What happens if I lower alpha from 0.05 to 0.01?

You make it harder to reject H₀, which cuts the chance of a Type I error to 1% but raises the chance of a Type II error (missing a real effect). That tradeoff between alpha and power shows up in Unit 6 multiple-choice questions.

## Related Study Guides

- [3.5 Setting Up a Test for a Population Proportion](/ap-stats/unit-3/setting-up-test-for-population-proportion/study-guide/QLu7hUN0rwtnxLF7YdBT)

## Structured Data

```json
{"@context":"https://schema.org","@graph":[{"@type":"LearningResource","@id":"https://fiveable.me/ap-stats/key-terms/significance-level-alpha#resource","name":"Significance Level (Alpha) — AP Stats Definition & Guide","url":"https://fiveable.me/ap-stats/key-terms/significance-level-alpha","learningResourceType":"Concept explainer","educationalLevel":"AP® / High School","about":{"@id":"https://fiveable.me/ap-stats/key-terms/significance-level-alpha#term"},"audience":{"@type":"EducationalAudience","educationalRole":"student"},"dateModified":"2026-06-11T05:22:54.437Z","isPartOf":{"@type":"Collection","name":"AP Statistics Key Terms","url":"https://fiveable.me/ap-stats/key-terms"},"publisher":{"@type":"Organization","name":"Fiveable","url":"https://fiveable.me"}},{"@type":"DefinedTerm","@id":"https://fiveable.me/ap-stats/key-terms/significance-level-alpha#term","name":"Significance level (alpha)","description":"The significance level (alpha, α) is the threshold probability, typically 0.05, set before a hypothesis test; if the p-value is less than α, you reject the null hypothesis. Alpha also equals the probability of making a Type I error (rejecting a true null hypothesis).","url":"https://fiveable.me/ap-stats/key-terms/significance-level-alpha","inDefinedTermSet":{"@type":"DefinedTermSet","name":"AP Statistics Key Terms","url":"https://fiveable.me/ap-stats/key-terms"}},{"@type":"FAQPage","mainEntity":[{"@type":"Question","name":"What is the significance level (alpha) in AP Stats?","acceptedAnswer":{"@type":"Answer","text":"It's the threshold probability, usually 0.05, that you set before a hypothesis test. If your p-value comes out below alpha, you reject the null hypothesis; otherwise you fail to reject it."}},{"@type":"Question","name":"Is alpha the same thing as the p-value?","acceptedAnswer":{"@type":"Answer","text":"No. Alpha is a fixed cutoff you choose before collecting data, while the p-value is calculated from your sample. The test decision comes from comparing the two, and confusing them is a frequent point-loser on FRQs."}},{"@type":"Question","name":"Does a result above alpha prove the null hypothesis is true?","acceptedAnswer":{"@type":"Answer","text":"No. When the p-value is at or above α, you fail to reject H₀, which only means the evidence wasn't strong enough. It never proves H₀ is correct, which is why \"accept H₀\" is wrong phrasing on the exam."}},{"@type":"Question","name":"Why is alpha usually 0.05?","acceptedAnswer":{"@type":"Answer","text":"It's a convention, not a law. α = 0.05 means tolerating a 5% chance of a Type I error, which balances catching real effects against false alarms. Problems use 0.01 when false alarms are costly and 0.10 when missing a real effect is worse."}},{"@type":"Question","name":"What happens if I lower alpha from 0.05 to 0.01?","acceptedAnswer":{"@type":"Answer","text":"You make it harder to reject H₀, which cuts the chance of a Type I error to 1% but raises the chance of a Type II error (missing a real effect). That tradeoff between alpha and power shows up in Unit 6 multiple-choice questions."}}]},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"AP Statistics","item":"https://fiveable.me/ap-stats"},{"@type":"ListItem","position":2,"name":"Key Terms","item":"https://fiveable.me/ap-stats/key-terms"},{"@type":"ListItem","position":3,"name":"Unit 3","item":"https://fiveable.me/ap-stats/unit-3"},{"@type":"ListItem","position":4,"name":"Significance level (alpha)"}]}]}
```