Hypothesis testing with the binomial distribution
A hypothesis test asks whether a result is surprising enough to doubt what you assumed. You assume the usual value of a probability, see how likely your data would be if that were true, and decide using a fixed cut-off.
Part 1 of 3: Learn it
In short
- H₀ states the parameter's assumed value; H₁ states how you suspect it has changed.
- If the probability of a result at least as extreme as yours is below the significance level, reject H₀.
- In a two-tailed test, split the significance level between the two tails.
Where this is in your specification
Spec points: DfE O1 and O2 (AQA 7357 O1 and O2, Edexcel 9MA0 Statistics topic 5, OCR A H240 2.05)
| Board | Topic: Hypothesis testing |
|---|---|
| DfE content | O1-O4 |
| AQA 7357 | O1-O4 |
| Edexcel 9MA0 | Statistics topic 5 |
| OCR H240 | 2.05 |
The language
| Term | Meaning |
|---|---|
| Null hypothesis H₀ | the assumption being tested, such as p = 0.5 |
| Alternative hypothesis H₁ | p > 0.5, p < 0.5 (one-tailed) or p ≠ 0.5 (two-tailed) |
| Significance level | the cut-off probability, often 5% or 1% |
| Test statistic | the value from the sample, such as the number of heads |
| Critical region | the values of the test statistic that would lead you to reject H₀ |
| Actual significance level | the true probability of landing in the critical region if H₀ is true |
The steps
- Define the parameter and write H₀ and H₁.
- Write the distribution of the test statistic assuming H₀ is true, such as X ~ B(20, 0.5).
- Find the probability of the observed value or anything more extreme in the direction of H₁.
- Compare it with the significance level (halved in each tail for a two-tailed test).
- Conclude in context, without claiming certainty.
Critical regions
Instead of a single probability, you can find every value that would cause H₀ to be rejected. List cumulative probabilities near the tail and choose the boundary where the tail probability first drops below the significance level. Because X is discrete, the actual significance level is usually a bit below the nominal one.
What does a two-tailed test at the 10% level put in each tail?
Show the answer
5% in each tail.
Part 2 of 3: See it worked
Worked examples
Example 1
A coin is suspected of landing heads more often than tails. In 20 flips it shows 15 heads. Test at the 5% level.
- Let p be the probability of heads. Test H₀: p = 0.5 against H₁: p > 0.5
- Under H₀, X ~ B(20, 0.5)
- P(X ≥ 15) = 1 − P(X ≤ 14) = 0.0207
- 0.0207 < 0.05, so reject H₀
Answer: There is sufficient evidence at the 5% level to suggest the coin is biased towards heads.
Example 2
For the same test, find the critical region and the actual significance level.
- P(X ≥ 14) = 0.0577, which is more than 0.05
- P(X ≥ 15) = 0.0207, which is less than 0.05
- So the critical region starts at 15
Answer: X ≥ 15, with an actual significance level of 2.07%.
Common mistakes
- Writing hypotheses about the sample (X) instead of the parameter (p).
- Finding P(X = 15) instead of P(X ≥ 15).
- Forgetting to halve the significance level for a two-tailed test.
- Concluding that H₀ has been proved true or false. A test only gives evidence.
In a test of H₀: p = 0.3 against H₁: p < 0.3, which values of X form the critical region: high or low?
Show the answer
Low values, in the lower tail.
Part 3 of 3: Test yourself
Check yourself
Answer each one in your head or on paper first, then open it to check.
What does a two-tailed test at the 10% level put in each tail?
5% in each tail.
In a test of H₀: p = 0.3 against H₁: p < 0.3, which values of X form the critical region: high or low?
Low values, in the lower tail.
Why is the actual significance level usually less than the stated one?
X only takes whole numbers, so the tail probabilities jump in steps and rarely hit the stated level exactly.
Jobs that use this
- Data scientist (s'ouvre dans un nouvel onglet)
- Data analyst-statistician (s'ouvre dans un nouvel onglet)
- Pharmacologist (s'ouvre dans un nouvel onglet)
Each link opens the job profile on the National Careers Service (England). In the rest of the UK: My World of Work (Scotland), Careers Wales, nidirect careers (Northern Ireland).
Ces fiches sont en anglais car elles suivent les programmes d'examen britanniques.
Full lessons and marked practice for this course are coming soon to Brainlag Learn. See courses