Scribe logoScribeNotes & practice
Sign inStart free
  1. Notes
  2. ›Edexcel A-level Mathematics
  3. ›S Statistics
  4. ›S5 Hypothesis testing

Edexcel A-level Mathematics · S · Statistics

S5 Hypothesis testing

Set up null and alternative hypotheses, carry out binomial hypothesis tests using significance levels and critical regions, and interpret the result in context.

  • 5 min read
  • 5 parts
  • 10 questions
  • Draft
Start a Quick 5 on this topic→Jump to the 10 questions
1/5A hypothesis test weighs a claim against sample evidence
  1. 1A hypothesis test weighs a claim against sample evidence
  2. 2The alternative hypothesis decides which tail the test uses
  3. 3The significance level sets how unlikely a result must be
  4. 4The critical region contains every result that rejects the null hypothesis
  5. 5Always state the conclusion in the context of the problem
  6. Check your recall
  7. Practice: 10 questions
1

Part 1 of 5: A hypothesis test weighs a claim against sample evidence

A friend says a coin is fair, then it lands heads 9 times in 10 spins. That result makes you doubt the claim. A hypothesis test puts this doubt on a firm footing.

The claim being tested is the null hypothesis, H0H_0H0​. It always states that a parameter takes one exact value. The alternative hypothesis, H1H_1H1​, says what the parameter may be instead.

For a binomial test the parameter is ppp, the probability of success on each trial. Write hypotheses in terms of ppp, never in terms of the sample.

Exam tip

Say what ppp means in the problem, for example ‘ppp is the probability that the coin lands heads’.

Key terms · tap one for its definition

hypothesis test

A hypothesis test is a procedure that uses sample data to decide whether there is enough evidence to reject a claim about a population parameter.

null hypothesis

The null hypothesis H0H_0H0​ is the default claim about a parameter, assumed true when calculating probabilities in the test.

alternative hypothesis

The alternative hypothesis H1H_1H1​ is the statement about the parameter that is supported if the null hypothesis is rejected.

Part 1 done

2

Part 2 of 5: The alternative hypothesis decides which tail the test uses

Now ask what you are looking for. The wording of the problem sets the form of H1H_1H1​. The null hypothesis always uses an equals sign: H0:p=p0H_0:p=p_0H0​:p=p0​. Choose the direction of H1H_1H1​ from the situation described, not from the sample result.

A one-tailed test looks for a change in one direction only, such as H1:p<p0H_1:p<p_0H1​:p<p0​. A two-tailed test looks for any change, so H1:p≠p0H_1:p\ne p_0H1​:p=p0​.

Choosing the alternative hypothesis

Wording of the problemH1H_1H1​Test
Fewer, lower, decreasedp<p0p<p_0p<p0​One-tailed, lower tail
More, higher, increasedp>p0p>p_0p>p0​One-tailed, upper tail
Different, biased, changedp≠p0p\ne p_0p=p0​Two-tailed
Wording of the problem
Fewer, lower, decreased
H1H_1H1​
p<p0p<p_0p<p0​
Test
One-tailed, lower tail
Wording of the problem
More, higher, increased
H1H_1H1​
p>p0p>p_0p>p0​
Test
One-tailed, upper tail
Wording of the problem
Different, biased, changed
H1H_1H1​
p≠p0p\ne p_0p=p0​
Test
Two-tailed

Common mistake

The word ‘biased’ gives no direction. Use a two-tailed test unless the problem says which way the bias goes.

Key terms · tap one for its definition

one-tailed test

A one-tailed test is a hypothesis test whose alternative hypothesis states that the parameter is greater than, or less than, the null value.

two-tailed test

A two-tailed test is a hypothesis test whose alternative hypothesis states only that the parameter differs from the null value.

Part 2 done

Quick check

One question on the part you just read

Easy

A manufacturer claims the probability that a bulb is faulty is 0.2. A buyer suspects it is higher. Identify the correct hypotheses.
Show answer and explanation

Answer: C · Worked explanation

The null hypothesis states one exact value, p=0.2p=0.2p=0.2. The buyer suspects the probability is higher, so H1:p>0.2H_1:p>0.2H1​:p>0.2. The version with p≠0.2p\ne0.2p=0.2 would be a two-tailed test, which does not match the buyer’s suspicion of an increase.
3

Part 3 of 5: The significance level sets how unlikely a result must be

You also need a rule for deciding when a result is too unlikely. The significance level is a probability, often 5%, chosen before the data are seen. A smaller significance level demands stronger evidence before H0H_0H0​ is rejected.

Assume H0H_0H0​ is true. The test statistic is the number of successes XXX in the sample, and it follows B(n,p0)B(n,p_0)B(n,p0​). Find the probability of a result at least as extreme as the one observed.

If this probability is no more than the significance level, reject H0H_0H0​. Otherwise do not reject H0H_0H0​.

Worked example

One-tailed test

A shop claims that 30% of visitors buy something. The owner thinks the true figure is lower. In a sample of 20 visitors, 2 buy. Test at the 5% level.
  1. 1
    Hypotheses: H0:p=0.3H_0:p=0.3H0​:p=0.3 and H1:p<0.3H_1:p<0.3H1​:p<0.3.
  2. 2
    Under H0H_0H0​, X∼B(20,0.3)X\sim B(20,0.3)X∼B(20,0.3).
  3. 3
    Calculate P(X≤2)=0.0355P(X\le2)=0.0355P(X≤2)=0.0355.
  4. 4
    Compare: 0.0355<0.050.0355<0.050.0355<0.05, so reject H0H_0H0​.

Answer: There is sufficient evidence at the 5% level that the proportion of visitors who buy is less than 0.3.

Key terms · tap one for its definition

significance level

The significance level is the probability, chosen in advance, at or below which a result is judged unlikely enough to reject the null hypothesis.

test statistic

The test statistic is the value calculated from the sample, such as the number of successes XXX, used to decide the outcome of the test.

Part 3 done

4

Part 4 of 5: The critical region contains every result that rejects the null hypothesis

Instead of testing one result, you can list every result that would lead to rejecting H0H_0H0​. That set of values is the critical region. Its boundary value is the critical value.

Take a lower-tail test with X∼B(10,0.5)X\sim B(10,0.5)X∼B(10,0.5) at 5%. Here P(X≤1)=0.0107P(X\le1)=0.0107P(X≤1)=0.0107 and P(X≤2)=0.0547P(X\le2)=0.0547P(X≤2)=0.0547. So the critical region is X≤1X\le1X≤1.

Figure 1. Values of X in the shaded left-hand boxes lead to rejecting the null hypothesis. All other values do not.

The probability of landing in the critical region when H0H_0H0​ is true is the actual significance level. Here it is 0.0107, or 1.07%, not 5%. In a two-tailed test, split the significance level equally between the tails. The critical region depends only on nnn, p0p_0p0​, the tail and the significance level, not on the data.

Key terms · tap one for its definition

critical region

The critical region is the set of values of the test statistic for which the null hypothesis is rejected.

critical value

The critical value is the value of the test statistic at the edge of the critical region, lying inside it.

actual significance level

The actual significance level is the probability that the test statistic falls in the critical region when the null hypothesis is true.

Part 4 done

Quick check

One question on the part you just read

Easy

Identify the alternative hypothesis that gives a two-tailed test.
Show answer and explanation

Answer: B · Worked explanation

A two-tailed test looks for any difference from the null value, so H1:p≠0.4H_1:p\ne0.4H1​:p=0.4. The statements with <<< and >>> each look in one direction only. The statement with === is the form of the null hypothesis.
5

Part 5 of 5: Always state the conclusion in the context of the problem

The last step turns the numbers back into words. A complete test follows four steps:

  • Write H0H_0H0​ and H1H_1H1​ in terms of ppp.
  • State the distribution of XXX under H0H_0H0​.
  • Compare the probability with the significance level.
  • Write a conclusion in the context of the problem.

If H0H_0H0​ is rejected, say there is sufficient evidence to suggest H1H_1H1​. If not, say there is insufficient evidence to reject H0H_0H0​. Name the real quantity, such as ‘the proportion of visitors who buy’.

Worked example

Two-tailed test

A coin is spun 10 times and lands heads 9 times. Test at the 5% level whether it is biased.
  1. 1
    Hypotheses: H0:p=0.5H_0:p=0.5H0​:p=0.5 and H1:p≠0.5H_1:p\ne0.5H1​:p=0.5.
  2. 2
    Under H0H_0H0​, X∼B(10,0.5)X\sim B(10,0.5)X∼B(10,0.5).
  3. 3
    Find P(X≥9)=1−P(X≤8)=1−0.9893=0.0107P(X\ge9)=1-P(X\le8)=1-0.9893=0.0107P(X≥9)=1−P(X≤8)=1−0.9893=0.0107.
  4. 4
    Each tail has 2.5%, so compare 0.0107 with 0.025. It is smaller, so reject H0H_0H0​.

Answer: There is sufficient evidence at the 5% level that the coin is biased.

Common mistake

Never write that H0H_0H0​ is proved true or that H1H_1H1​ is certain. A test gives evidence, not proof.

Note read

Check your recall[13 marks]

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

  1. 1.State[2 marks]

    State suitable hypotheses for a test of whether a coin is biased towards heads, where ppp is the probability of heads.

    Show mark schemeHide mark scheme

    Model answer

    H0:p=0.5H_0:p=0.5H0​:p=0.5 and H1:p>0.5H_1:p>0.5H1​:p>0.5.

    • 1 mark for stating H0:p=0.5H_0:p=0.5H0​:p=0.5
    • 1 mark for stating H1:p>0.5H_1:p>0.5H1​:p>0.5
  2. 2.Explain[2 marks]

    Explain the difference between a one-tailed test and a two-tailed test.

    Show mark schemeHide mark scheme

    Model answer

    A one-tailed test has an alternative hypothesis in one direction only, such as p<p0p<p_0p<p0​. A two-tailed test has p≠p0p\ne p_0p=p0​, so it looks for a change in either direction.

    • 1 mark for stating that a one-tailed test looks for a change in one direction
    • 1 mark for stating that a two-tailed test looks for a change in either direction
  3. 3.Define[2 marks]

    Define the term critical region.

    Show mark schemeHide mark scheme

    Model answer

    The critical region is the set of values of the test statistic for which the null hypothesis is rejected.

    • 1 mark for referring to values of the test statistic
    • 1 mark for stating that these values lead to rejecting the null hypothesis
  4. 4.Determine[3 marks]

    Determine whether H0:p=0.5H_0:p=0.5H0​:p=0.5 is rejected in favour of H1:p>0.5H_1:p>0.5H1​:p>0.5 at the 5% level, when X∼B(10,0.5)X\sim B(10,0.5)X∼B(10,0.5) gives X=9X=9X=9 and P(X≥9)=0.0107P(X\ge9)=0.0107P(X≥9)=0.0107.

    Show mark schemeHide mark scheme

    Model answer

    Compare 0.0107 with 0.05. Since 0.0107<0.050.0107<0.050.0107<0.05, reject H0H_0H0​. There is sufficient evidence at the 5% level that ppp is greater than 0.5.

    • 1 mark for comparing 0.0107 with the 5% significance level
    • 1 mark for rejecting H0H_0H0​
    • 1 mark for stating a conclusion in context that ppp is greater than 0.5
  5. 5.Explain[2 marks]

    Explain why a result that does not reject H0H_0H0​ does not prove that H0H_0H0​ is true.

    Show mark schemeHide mark scheme

    Model answer

    The test only shows that the sample gives insufficient evidence against H0H_0H0​. The true value of ppp may still differ from the claim.

    • 1 mark for stating that there is insufficient evidence against H0H_0H0​
    • 1 mark for stating that ppp may still differ from the claimed value
  6. 6.Explain[2 marks]

    Explain what is meant by a significance level of 5%.

    Show mark schemeHide mark scheme

    Model answer

    It is the probability threshold chosen in advance. If the probability of a result at least as extreme as the observed one is 5% or less under H0H_0H0​, then H0H_0H0​ is rejected.

    • 1 mark for stating that it is a probability threshold chosen in advance
    • 1 mark for stating that H0H_0H0​ is rejected when the result’s probability under H0H_0H0​ is at or below it

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 what a significance level of 5% means in a hypothesis test.
Show answer and explanation

Answer: D · Worked explanation

The significance level is a probability threshold chosen in advance. When the probability of a result at least as extreme is at or below it, H0H_0H0​ is rejected. It is not the probability that H0H_0H0​ is true, because H0H_0H0​ is assumed true when calculating.
4.

Medium·Determine[1 mark]

X∼B(20,0.3)X\sim B(20,0.3)X∼B(20,0.3) under H0:p=0.3H_0:p=0.3H0​:p=0.3, with H1:p<0.3H_1:p<0.3H1​:p<0.3 tested at 5%. The observed value is X=3X=3X=3 and P(X≤3)=0.1071P(X\le3)=0.1071P(X≤3)=0.1071. Determine the correct conclusion.
Show answer and explanation

Answer: A · Worked explanation

Compare 0.1071 with the significance level 0.05. Since 0.1071>0.050.1071>0.050.1071>0.05, the result is not unlikely enough, so H0H_0H0​ is not rejected. Rejecting because 3 is below the expected 6 ignores the significance level, since a value below average is not enough by itself.
5.

Medium·Determine[2 marks]

X∼B(10,0.5)X\sim B(10,0.5)X∼B(10,0.5) under H0:p=0.5H_0:p=0.5H0​:p=0.5, with H1:p<0.5H_1:p<0.5H1​:p<0.5. Given P(X≤1)=0.0107P(X\le1)=0.0107P(X≤1)=0.0107, P(X≤2)=0.0547P(X\le2)=0.0547P(X≤2)=0.0547 and P(X≤3)=0.1719P(X\le3)=0.1719P(X≤3)=0.1719, determine the critical region at the 5% level.
Show answer and explanation

Answer: C · Worked explanation

The critical region is the largest set of low values whose total probability is at most 0.05. Since P(X≤1)=0.0107≤0.05P(X\le1)=0.0107\le0.05P(X≤1)=0.0107≤0.05 but P(X≤2)=0.0547>0.05P(X\le2)=0.0547>0.05P(X≤2)=0.0547>0.05, the region is X≤1X\le1X≤1. The region X≤2X\le2X≤2 looks closest to 5% but exceeds it.
6.

Medium·Determine[2 marks]

X∼B(10,0.5)X\sim B(10,0.5)X∼B(10,0.5) under H0:p=0.5H_0:p=0.5H0​:p=0.5 with H1:p≠0.5H_1:p\ne0.5H1​:p=0.5 at the 10% level. The observed value is X=8X=8X=8 and P(X≥8)=0.0547P(X\ge8)=0.0547P(X≥8)=0.0547. Determine the correct conclusion.
Show answer and explanation

Answer: B · Worked explanation

A two-tailed test at 10% puts 5% in each tail, so compare 0.0547 with 0.05. Since 0.0547>0.050.0547>0.050.0547>0.05, do not reject H0H_0H0​. Comparing with 0.10 would be a mistake, because that figure is the total for both tails rather than the figure for one tail.
7.

Medium·Determine[2 marks]

X∼B(10,0.5)X\sim B(10,0.5)X∼B(10,0.5) under H0:p=0.5H_0:p=0.5H0​:p=0.5 with H1:p≠0.5H_1:p\ne0.5H1​:p=0.5 at the 5% level. Given P(X≤1)=0.0107P(X\le1)=0.0107P(X≤1)=0.0107, P(X≤2)=0.0547P(X\le2)=0.0547P(X≤2)=0.0547, P(X≥8)=0.0547P(X\ge8)=0.0547P(X≥8)=0.0547 and P(X≥9)=0.0107P(X\ge9)=0.0107P(X≥9)=0.0107, determine the critical region.
Show answer and explanation

Answer: D · Worked explanation

Each tail may have at most 2.5%, which is 0.025. The probability 0.0107 is below 0.025 but 0.0547 is above it. So the critical region is X≤1X\le1X≤1 or X≥9X\ge9X≥9. The region X≤2X\le2X≤2 or X≥8X\ge8X≥8 would give each tail 0.0547, which is too large.
8.

Hard·Determine[2 marks]

A supplier claims that 80% of seeds germinate. A gardener plants 20 seeds and 12 germinate. Under H0:p=0.8H_0:p=0.8H0​:p=0.8 with H1:p<0.8H_1:p<0.8H1​:p<0.8, P(X≤12)=0.0321P(X\le12)=0.0321P(X≤12)=0.0321. Determine the best conclusion at the 5% level.
Show answer and explanation

Answer: A · Worked explanation

Since 0.0321<0.050.0321<0.050.0321<0.05, H0H_0H0​ is rejected, and the conclusion is stated as evidence in context. The value 0.0321 is not the probability that the claim is true, because the test assumes H0H_0H0​. Evidence from one sample does not prove anything about every packet.
9.

Hard·Calculate[2 marks]

A test uses the critical region X≤1X\le1X≤1 for X∼B(10,0.5)X\sim B(10,0.5)X∼B(10,0.5) under H0:p=0.5H_0:p=0.5H0​:p=0.5, with P(X≤1)=0.0107P(X\le1)=0.0107P(X≤1)=0.0107. Calculate the probability that H0H_0H0​ is rejected when H0H_0H0​ is true.
Show answer and explanation

Answer: B · Worked explanation

When H0H_0H0​ is true, X∼B(10,0.5)X\sim B(10,0.5)X∼B(10,0.5), so H0H_0H0​ is rejected exactly when X≤1X\le1X≤1. That probability is the actual significance level, 0.0107. The value 0.05 is the chosen significance level, not the probability for this critical region.
10.

Hard·Determine[2 marks]

An observed result has a lower-tail probability of 0.04 under H0H_0H0​. Determine which statement is correct at the 5% significance level.
Show answer and explanation

Answer: C · Worked explanation

In a one-tailed test, compare 0.04 with 0.05. It is smaller, so reject H0H_0H0​. In a two-tailed test, each tail has 0.025 and 0.04>0.0250.04>0.0250.04>0.025, so do not reject H0H_0H0​. A two-tailed test is harder to reject, not easier.

Keep going

Next topicM6 Quantities and units in mechanicsPrevious topicS4 Statistical distributions

More in S · Statistics

  1. S1Statistical samplingNotes · 10 questions
  2. S2Data presentation and interpretationNotes · 10 questions
  3. S3ProbabilityNotes · 10 questions
  4. S4Statistical distributionsNotes · 10 questions
  5. S5Hypothesis testingNotes · 10 questions

All Edexcel A-level Mathematics 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 Pearson Edexcel or any exam board.

On this page

0/5 read

  1. 1A hypothesis test weighs a claim against sample evidence
  2. 2The alternative hypothesis decides which tail the test uses
  3. 3The significance level sets how unlikely a result must be
  4. 4The critical region contains every result that rejects the null hypothesis
  5. 5Always state the conclusion in the context of the problem
  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