Scribe logoScribeNotes & practice
Sign inStart free
  1. Notes
  2. ›AP Statistics
  3. ›3 Inference for Categorical Data: Proportions
  4. ›3.7 Carrying out a test for a proportion

AP Statistics · Unit 3 · Inference for Categorical Data: Proportions

3.7 Carrying out a test for a proportion

Carry out a test for a proportion, decide whether to reject the null hypothesis, and state the conclusion in context.

  • 4 min read
  • 5 parts
  • 10 questions
  • Draft
Start a Quick 5 on this topic→Jump to the 10 questions
1/5A test follows four steps from hypotheses to conclusion
  1. 1A test follows four steps from hypotheses to conclusion
  2. 2The z statistic measures distance from the null value in standard deviations
  3. 3Comparing the p-value with the significance level gives the decision
  4. 4A conclusion in context links the decision to the population proportion
  5. 5A decision gives evidence about the population, not proof
  6. Check your recall
  7. Practice: 10 questions
1

Part 1 of 5: A test follows four steps from hypotheses to conclusion

You can now set up a test and interpret a p-value. Next, combine those skills into one routine. Every test of a proportion follows the same four steps. Keeping them in order prevents most careless errors.

  • State the hypotheses, defining ppp in context.
  • Check the conditions for a normal model.
  • Calculate the test statistic and the p-value.
  • Compare the p-value with α\alphaα, then conclude in context.

A test statistic turns the sample result into one number. For a proportion it is a z-score, which counts standard deviations from the null value.

Key term

test statistic

A test statistic is a value computed from sample data that measures how far the result lies from the null hypothesis value.

Part 1 done

2

Part 2 of 5: The z statistic measures distance from the null value in standard deviations

Start with the idea. The difference between the sample proportion and the null value shows how far the result is from the claim. Dividing by the typical size of sampling variation standardizes that difference.

That gives z=p^−p0p0(1−p0)nz=\frac{\hat{p}-p_0}{\sqrt{\frac{p_0(1-p_0)}{n}}}z=np0​(1−p0​)​​p^​−p0​​. The denominator uses p0p_0p0​, because the test assumes the null hypothesis is true. A z value near 0 means the result is close to the claim. A z value far from 0 means it is not.

The p-value is the area beyond zzz in the direction of HaH_aHa​. For Ha:p≠p0H_a: p \neq p_0Ha​:p=p0​, double the one-tail area.

Worked example

Finding z and the p-value

A company claims that 30% of its 5,000 customers renew each year. In a random sample of 200 customers, 48 renewed. Test H0:p=0.30H_0: p = 0.30H0​:p=0.30 against Ha:p<0.30H_a: p < 0.30Ha​:p<0.30. Find z and the p-value.
  1. 1
    Conditions: the sample is random, 200 is less than 10% of 5,000, and 200(0.30)=60200(0.30) = 60200(0.30)=60 and 200(0.70)=140200(0.70) = 140200(0.70)=140 are at least 10.
  2. 2
    Sample proportion: p^=48200=0.24\hat{p}=\frac{48}{200}=0.24p^​=20048​=0.24.
  3. 3
    Standard deviation: 0.30(0.70)200≈0.0324\sqrt{\frac{0.30(0.70)}{200}} \approx 0.03242000.30(0.70)​​≈0.0324.
  4. 4
    Test statistic: z=0.24−0.300.0324≈−1.85z=\frac{0.24-0.30}{0.0324} \approx -1.85z=0.03240.24−0.30​≈−1.85.
  5. 5
    The alternative is lower, so the p-value is the lower-tail area below -1.85, which is about 0.032.

Answer: z≈−1.85z \approx -1.85z≈−1.85 and the p-value is about 0.032.

Part 2 done

Quick check

One question on the part you just read

Easy

A test of a proportion gives a p-value of 0.01, and α=0.05\alpha = 0.05α=0.05. Identify the correct decision.
Show answer and explanation

Answer: C · Worked explanation

The p-value of 0.01 is at or below 0.05, so the decision is to reject the null hypothesis. Failing to reject would be correct only when the p-value is above α\alphaα. The alternative is never rejected or accepted in this way.
3

Part 3 of 5: Comparing the p-value with the significance level gives the decision

Now make the decision. The significance level α\alphaα is the cutoff chosen before the data are seen. The rule has two outcomes, and each has its own wording.

  • If the p-value is at or below α\alphaα, reject the null hypothesis. The data give convincing evidence for the alternative.
  • If the p-value is above α\alphaα, fail to reject the null hypothesis. The data do not give convincing evidence for the alternative.

In the renewal example, 0.032 is below α=0.05\alpha = 0.05α=0.05, so we reject H0H_0H0​.

Common mistake

Do not say accept the null hypothesis. A large p-value means the data fail to show the null is wrong, not that it is true.

Key terms · tap one for its definition

reject the null hypothesis

To reject the null hypothesis is to decide that the p-value is at or below α\alphaα and the data give convincing evidence against it.

fail to reject the null hypothesis

To fail to reject the null hypothesis is to decide that the p-value is above α\alphaα, so the data do not give convincing evidence against it.

Part 3 done

4

Part 4 of 5: A conclusion in context links the decision to the population proportion

Next, write the conclusion. It connects the p-value, the decision, and the population. A complete answer has three parts.

  • Compare: the p-value of 0.032 is less than α=0.05\alpha = 0.05α=0.05.
  • Decide: we reject H0H_0H0​.
  • Conclude: there is convincing evidence that the proportion of all customers who renew is less than 0.30.

A conclusion about the sample alone earns no credit. The claim concerns the population, so the sentence must name it.

Exam tip

Name the population, the proportion, and the direction in HaH_aHa​. Link the p-value, α\alphaα, the decision, and the context in one statement.

Part 4 done

Quick check

One question on the part you just read

Easy

Which phrase correctly names the decision when the p-value is greater than α\alphaα?
Show answer and explanation

Answer: A · Worked explanation

A p-value above α\alphaα means the data do not give convincing evidence against the null, so we fail to reject it. Accepting or proving the null overstates what a large p-value shows. Rejecting the null requires a p-value at or below α\alphaα.
5

Part 5 of 5: A decision gives evidence about the population, not proof

Finally, consider what each decision means. Neither one proves anything, and the wording must reflect that.

What each decision means

DecisionWhat it meansWhat it does not mean
Reject H0H_0H0​Data are unlikely under the null; convincing evidence for HaH_aHa​HaH_aHa​ is proven, or the null is impossible
Fail to reject H0H_0H0​Data are consistent with the null; no convincing evidence for HaH_aHa​The null is true, or ppp equals p0p_0p0​
Decision
Reject H0H_0H0​
What it means
Data are unlikely under the null; convincing evidence for HaH_aHa​
What it does not mean
HaH_aHa​ is proven, or the null is impossible
Decision
Fail to reject H0H_0H0​
What it means
Data are consistent with the null; no convincing evidence for HaH_aHa​
What it does not mean
The null is true, or ppp equals p0p_0p0​

When you reject H0H_0H0​, a small chance remains that random sampling misled you. When you fail to reject, the null value is only one plausible value among many. A random sample lets you carry either conclusion to the whole population.

Suppose the p-value is 0.20 for Ha:p≠0.50H_a: p \neq 0.50Ha​:p=0.50. The correct statement is that there is not convincing evidence that the proportion differs from 0.50. It is not a statement that the proportion equals 0.50.

Note read

Check your recall[12 marks]

5 written questions. Answer each one in your head or on paper, then open the mark scheme.

  1. 1.Calculate[3 marks]

    A random sample of 150 customers, fewer than 10% of all customers, has 57 who prefer a new product. Calculate the z statistic for testing H0:p=0.40H_0: p = 0.40H0​:p=0.40.

    Show mark schemeHide mark scheme

    Model answer

    p^=57150=0.38\hat{p}=\frac{57}{150}=0.38p^​=15057​=0.38. The standard deviation is 0.40(0.60)150=0.04\sqrt{\frac{0.40(0.60)}{150}}=0.041500.40(0.60)​​=0.04. Then z=0.38−0.400.04=−0.5z=\frac{0.38-0.40}{0.04}=-0.5z=0.040.38−0.40​=−0.5.

    • 1 mark for calculating p^=0.38\hat{p}=0.38p^​=0.38
    • 1 mark for calculating the standard deviation 0.04 using p0=0.40p_0 = 0.40p0​=0.40
    • 1 mark for obtaining z=−0.5z=-0.5z=−0.5
  2. 2.Determine[2 marks]

    Determine whether to reject H0H_0H0​ when the p-value is 0.032 and α=0.01\alpha = 0.01α=0.01.

    Show mark schemeHide mark scheme

    Model answer

    Do not reject. The p-value of 0.032 is greater than 0.01, so we fail to reject H0H_0H0​.

    • 1 mark for comparing 0.032 with 0.01
    • 1 mark for concluding that we fail to reject H0H_0H0​
  3. 3.Justify[3 marks]

    A test of H0:p=0.50H_0: p = 0.50H0​:p=0.50 against Ha:p≠0.50H_a: p \neq 0.50Ha​:p=0.50 about the proportion of all voters in a county who support a measure gives a p-value of 0.20, with α=0.05\alpha = 0.05α=0.05. Justify the decision and give the conclusion in context.

    Show mark schemeHide mark scheme

    Model answer

    The p-value of 0.20 is greater than 0.05, so we fail to reject H0H_0H0​. There is not convincing evidence that the proportion of all voters in the county who support the measure differs from 0.50.

    • 1 mark for comparing 0.20 with 0.05 and deciding to fail to reject H0H_0H0​
    • 1 mark for stating that there is not convincing evidence for the alternative
    • 1 mark for giving the conclusion about the proportion of all voters in the county
  4. 4.Explain[2 marks]

    Explain why failing to reject the null hypothesis does not show that the null hypothesis is true.

    Show mark schemeHide mark scheme

    Model answer

    A large p-value means the data are consistent with the null value but do not prove it. Many other values of the proportion could also be consistent with the data.

    • 1 mark for stating that the data are consistent with the null but do not prove it
    • 1 mark for noting that other values of the proportion could also fit the data
  5. 5.Describe[2 marks]

    Describe what it means to reject the null hypothesis.

    Show mark schemeHide mark scheme

    Model answer

    The p-value is at or below α\alphaα. The data give convincing evidence against the null hypothesis and for the alternative.

    • 1 mark for stating that the p-value is at or below α\alphaα
    • 1 mark for stating that there is convincing evidence for the alternative

Practice this topic

10 quick-fire multiple-choice questions, easy to hard. Tap an answer to see if you're right and the worked explanation. (The written questions with mark schemes are under “Check your recall” above.)

0 right0 of 10 answered

Includes the 2 quick checks in the note above.

3.

Easy·Identify[1 mark]

Identify the formula for the z statistic in a test of a proportion.
Show answer and explanation

Answer: D · Worked explanation

The numerator is the sample proportion minus the null value, and the denominator is the standard deviation of the sampling variation under the null, so it uses p0p_0p0​ and nnn. Using p^\hat{p}p^​ in the denominator is the interval formula. The other two forms omit the square root or the sample size.
4.

Medium[1 mark]

A random sample of 400 gives a sample proportion of 0.63. Calculate z for H0:p=0.60H_0: p = 0.60H0​:p=0.60.
Show answer and explanation

Answer: A · Worked explanation

The standard deviation is 0.60(0.40)400≈0.0245\sqrt{\frac{0.60(0.40)}{400}} \approx 0.02454000.60(0.40)​​≈0.0245, so z=0.63−0.600.0245≈1.22z=\frac{0.63-0.60}{0.0245} \approx 1.22z=0.02450.63−0.60​≈1.22. Using p^\hat{p}p^​ in the denominator gives about 1.24. Leaving out the division gives 0.03, and forgetting the square root gives about 50. The correct answer is the only one that uses p0p_0p0​ and the full standard deviation.
5.

Medium[1 mark]

A test of H0:p=0.90H_0: p = 0.90H0​:p=0.90 against Ha:p<0.90H_a: p < 0.90Ha​:p<0.90 about the proportion of all students who own a laptop gives a p-value of 0.021, with α=0.05\alpha = 0.05α=0.05. Which conclusion is best?
Show answer and explanation

Answer: B · Worked explanation

The p-value of 0.021 is below 0.05, so we reject the null and conclude about the population proportion in the direction of the alternative. A conclusion that the proportion equals 0.90 contradicts the decision, and a statement about the sample alone does not address the claim.
6.

Medium[1 mark]

A test of H0:p=0.25H_0: p = 0.25H0​:p=0.25 against Ha:p≠0.25H_a: p \neq 0.25Ha​:p=0.25 gives a p-value of 0.18, with α=0.05\alpha = 0.05α=0.05. Which conclusion is correct?
Show answer and explanation

Answer: A · Worked explanation

The p-value of 0.18 is greater than 0.05, so we fail to reject the null. We can say the data are not convincing evidence of a difference, not that the proportion is 0.25. Rejecting would need a p-value at or below 0.05.
7.

Medium[1 mark]

A test gives a p-value of 0.03. Predict what happens to the decision if α\alphaα is changed from 0.05 to 0.01.
Show answer and explanation

Answer: C · Worked explanation

At α=0.05\alpha = 0.05α=0.05, the p-value of 0.03 is below the cutoff, so we reject. At α=0.01\alpha = 0.01α=0.01, the same p-value is above the cutoff, so we fail to reject. The p-value is calculated from the data, so changing α\alphaα does not change it.
8.

Hard[2 marks]

A test of H0:p=0.50H_0: p = 0.50H0​:p=0.50 with n=60n = 60n=60 and a sample proportion of 0.55 fails to reject H0H_0H0​. Which statement is most appropriate?
Show answer and explanation

Answer: A · Worked explanation

Failing to reject means the null value stays plausible, but it does not single out 0.50. Values around the sample proportion, 0.55, are also consistent with the data. A sample of 60 can reject a null value when the difference is large enough.
9.

Hard[2 marks]

After a test, a student writes, “We reject H0H_0H0​ and conclude that the sample proportion is greater than 0.40.” Which critique is best?
Show answer and explanation

Answer: D · Worked explanation

Tests judge claims about the population, so the conclusion must describe the population proportion in context. The sample proportion is already known, so concluding something about it adds nothing. Rejection gives evidence, not proof, and the null value is not the sample result.
10.

Hard[2 marks]

A random sample of 90, from a very large population, has a sample proportion of 0.40. Test H0:p=0.30H_0: p = 0.30H0​:p=0.30 against Ha:p>0.30H_a: p > 0.30Ha​:p>0.30 at α=0.02\alpha = 0.02α=0.02. Determine the correct result.
Show answer and explanation

Answer: B · Worked explanation

The standard deviation is 0.30(0.70)90≈0.0483\sqrt{\frac{0.30(0.70)}{90}} \approx 0.0483900.30(0.70)​​≈0.0483, so z=0.100.0483≈2.07z=\frac{0.10}{0.0483} \approx 2.07z=0.04830.10​≈2.07. The alternative is one-sided, so the p-value is about 0.019, which is below 0.02, and we reject. Doubling gives 0.038 and wrongly fails to reject. Skipping standardization gives z of 0.10.

Keep going

Next topic3.8 Errors in testsPrevious topic3.6 p-values

More in Unit 3 · Inference for Categorical Data: Proportions

  1. 3.3Confidence interval for a proportionNotes · 10 questions
  2. 3.4Claims from a proportion intervalNotes · 10 questions
  3. 3.5Setting up a test for a proportionNotes · 10 questions
  4. 3.6p-valuesNotes · 10 questions
  5. 3.7Carrying out a test for a proportionNotes · 10 questions
  6. 3.8Errors in testsNotes · 10 questions
  7. 3.9Differences in sample proportionsNotes · 10 questions
  8. 3.10Interval for a difference of proportionsNotes · 10 questions
  9. 3.11Claims from a difference intervalNotes · 10 questions

All AP Statistics topics →

Draft: written by Scribe and checked automatically, not yet by a subject reviewer.

Practice questions are original, written by Scribe and checked automatically for correctness.

Topic codes follow the official numbering. Not endorsed by College Board or any exam board.

On this page

0/5 read

  1. 1A test follows four steps from hypotheses to conclusion
  2. 2The z statistic measures distance from the null value in standard deviations
  3. 3Comparing the p-value with the significance level gives the decision
  4. 4A conclusion in context links the decision to the population proportion
  5. 5A decision gives evidence about the population, not proof
  6. Check your recall
  7. Practice: 10 questions
All coursesScribeHow it worksPricingPrivacyTerms

Notes and questions are written by Scribe in our own words. Topic codes follow the official course numbering so you can find your place. Not produced or endorsed by the IB, College Board, AQA, OCR, Pearson Edexcel or any exam board.

© 2026 Scribe