P-Value Examples: 46 Problems with Answers

Worked examples of P-Value to try before you open each solution, then practice problems with answers.

Worked Examples

Example 1

medium
A hypothesis test produces z=2.3 for a two-tailed test. Calculate the p-value and interpret it at both α=0.05 and α=0.01.
Show solution

Answer

p=0.0214. Significant at α=0.05 but not at α=0.01.

First step

1.
Two-tailed p-value: p=2×P(Z>2.3)=2×(1−0.9893)=2×0.0107=0.0214

See the full worked solution + why-it-works coaching

SetupKey insightWhy it worksCommon pitfallConnection

Free: full solutions on 3 topics a day. No card.

Every topic, every day: the Family plan

Example 2

hard
Correct the following misconceptions about p-values: (a) 'p=0.03 means there's a 3% chance H₀ is true.' (b) 'p=0.03 means the effect is large.'

Example 3

easy
A right-tailed test produces z=0.5. Use P(Z>0.5)=0.3085 to find the p-value and decide at α=0.05.

Example 4

medium
A one-sample z-test has xˉ=52, μ0=50, σ=8, n=64, with Ha:μ>50. Compute z and the p-value using P(Z>2)=0.0228.

Example 5

medium
A coin is flipped 100 times and lands heads 60 times. Test H0:p=0.5 vs Ha:p≠0.5 using a z-test. Use P(Z>2)=0.0228.

Example 6

medium
A two-tailed test produces z=2.58 with P(Z>2.58)=0.0049. Find the p-value and decide at α=0.01.

Example 7

hard
A t-test with df=9 produces t=2.262, the critical value for α=0.05 two-sided. Find the p-value.

Example 8

hard
Show that the probability of getting at least one p<0.05 in k=10 independent tests, all under H0, is 1−0.9510. Compute the value.

Example 9

hard
Explain the difference between P(data∣H0) (the p-value) and P(H0∣data) (the posterior on H0). Why are they not equal?

Example 10

hard
A test gives z=1.0 with n=25. If n were quadrupled to 100 with the same xˉ−μ0 and σ, what would z become, and how does the p-value change?

Example 11

challenge
Prove: if the test statistic T has a continuous distribution under H0, then the p-value P=P(T≥Tobs∣H0) is uniformly distributed on (0,1) when H0 holds.

Practice Problems

Try each one, then open the solution to compare your method.

Example 1

easy
A one-tailed test has z=1.7. Find the p-value and determine if we reject H0 at α=0.05.

Example 2

hard
Study A: n=50, p=0.04, effect size d=0.15 (tiny). Study B: n=10, p=0.06, effect size d=0.8 (large). Discuss which study's result is more practically meaningful and why we shouldn't rely solely on p-values.

Example 3

easy
Define in words what a p-value measures.

Example 4

easy
A test gives p=0.03 with α=0.05. Reject or fail to reject H0?

Example 5

easy
A test gives p=0.20. Does this prove the null hypothesis is true?

Example 6

easy
True or false: the p-value is the probability that H0 is true.

Example 7

easy
Which p-value is stronger evidence against H0: 0.001 or 0.04?

Example 8

easy
For a right-tailed test with z=1.645, the p-value equals which tail area? Use P(Z>1.645)=0.05.

Example 9

easy
Is α=0.05 a fundamental law of nature or a convention?

Example 10

easy
A two-sided test has one-tail area 0.015. What is the p-value?

Example 11

medium
A right-tailed test has z=2.33 with P(Z>2.33)≈0.0099. With α=0.01, conclude.

Example 12

medium
A two-sided test has z=−2.5 with P(Z<−2.5)≈0.0062. Find the p-value.

Example 13

medium
A study with xˉ=105, μ0=100, σ=15, n=9 tests Ha:μ≠100. Find z, then the p-value uses P(Z>1)=0.1587.

Example 14

medium
Two studies report p=0.049 and p=0.051 at α=0.05. How different is the underlying evidence?

Example 15

medium
A test gives p=0.30. A student writes 'there is a 30% chance H0 is true.' Correct the interpretation.

Example 16

medium
A left-tailed test Ha:μ<50 has z=−1.28 with P(Z<−1.28)≈0.10. With α=0.05, conclude.

Example 17

medium
If H0 is true, what is the distribution of the p-value (for a continuous test statistic)?

Example 18

challenge
Show why, under H0, P(p-value≤α)=α, and connect this to the Type I error rate.

Example 19

challenge
A journal publishes only p<0.05 results. Explain why published effect sizes tend to be overstated (the 'winner's curse').

Example 20

challenge
A test statistic has p-value 0.04. The researcher then collects more data hoping to 'confirm' significance, stopping when p<0.05 again. Why does this invalidate the p-value?

Example 21

medium
A right-tailed test has z=1.96 with P(Z>1.96)=0.025. Conclude at α=0.05.

Example 22

medium
A two-sided test has z=3 with P(Z>3)≈0.0013. Find the p-value.

Example 23

easy
A right-tailed z-test gives z=1.28. Using P(Z>1.28)=0.10, state the p-value.

Example 24

easy
A two-sided test produces one-tail area 0.025. Find the p-value.

Example 25

easy
With α=0.10 and p=0.07, do we reject H0?

Example 26

easy
Order from strongest to weakest evidence against H0: p=0.4, p=0.001, p=0.05.

Example 27

medium
A two-sided test produces z=−1.96 with P(Z<−1.96)=0.025. Find the p-value.

Example 28

medium
A study reports p=0.045. A reader claims 'there is a 4.5% chance the treatment has no effect.' Why is this wrong?

Example 29

medium
A two-sided test has z=1.5 with P(Z>1.5)=0.0668. Find the p-value.

Example 30

medium
Two studies test the same drug. Study A: n=1000, p=0.001, effect tiny. Study B: n=20, p=0.20, effect large. Which has more practical significance?

Example 31

medium
What value of α corresponds to a 99% confidence level?

Example 32

hard
A researcher runs 20 independent tests, all with true H0, at α=0.05 each. Expected number of false positives?

Example 33

hard
Bonferroni correction for k tests sets the per-test threshold at α/k. For α=0.05 family-wise and k=10 tests, what is the per-test threshold?

Example 34

hard
A study reports 'no significant difference, p=0.32.' Can we conclude there is no difference between groups?

Example 35

hard
Explain why p-hacking (running many tests until something is significant) inflates the actual Type I error far above the nominal α.

What to focus on

Core idea: The p-value is the probability of getting data at least this extreme assuming the null hypothesis is true.

Common stuck point: The procedure for p-value is the easy part; the trap is reading the p-value as the probability the null hypothesis is true. Asking "Am I computing the probability of data this extreme assuming the null is true (not the probability the null is true)?" first is what keeps a correct-looking calculation from being attached to the wrong concept.

Ask yourself: Am I computing the probability of data this extreme assuming the null is true (not the probability the null is true)?

Concept recap

The probability of observing a test statistic at least as extreme as the one computed from the sample data, assuming the null hypothesis H0 is true.

A p-value is a surprise meter that assumes nothing special is going on. A friend says she is an average free-throw shooter, then sinks 20 shots in a row; you ask how often an average shooter would do that. The rarer that is under her story, the less you believe the story. The meter rates the data, not the story's probability.

Read the full lesson on P-Value

Background knowledge

These ideas may be useful before you work through the harder examples.

hypothesis testingprobability