What is a hypothesis test in the context of Cambridge International A Levels Mathematics?

A hypothesis test in Cambridge International A Levels Mathematics, specifically in Probability and Statistics 2, is a statistical method used to determine whether there is enough evidence to reject a null hypothesis. It involves comparing sample data against a population parameter to make inferences about the population.

How do you determine the null and alternative hypotheses in a hypothesis test?

In a hypothesis test, the null hypothesis (H0) is a statement that there is no effect or no difference, and it serves as the default assumption. The alternative hypothesis (H1) is what you aim to support, indicating the presence of an effect or difference. For example, if testing a new drug, H0 might state that the drug has no effect, while H1 suggests it does have an effect.

What is the significance level in hypothesis testing, and why is it important?

The significance level, often denoted by alpha (α), is the probability of rejecting the null hypothesis when it is true. It is a threshold set by the researcher, commonly at 0.05, which determines the level of risk they are willing to take for a Type I error. It is crucial because it influences the decision-making process in hypothesis testing.

Can you explain the difference between Type I and Type II errors in hypothesis testing?

In hypothesis testing, a Type I error occurs when the null hypothesis is rejected when it is actually true, while a Type II error happens when the null hypothesis is not rejected when it is false. Understanding these errors is essential for interpreting the results of a hypothesis test accurately.

What role does the p-value play in hypothesis testing?

The p-value in hypothesis testing measures the probability of obtaining results at least as extreme as the observed results, assuming the null hypothesis is true. A small p-value (typically ≤ 0.05) indicates strong evidence against the null hypothesis, leading to its rejection. It helps in determining the statistical significance of the test results.

Why is "Wrong rejection region for one-tailed test" a common mistake in Hypothesis tests?

Why it happens: Looking at wrong tail. How to avoid it: H_1: > \mu_0 → reject in UPPER tail. H_1: < \mu_0 → reject in LOWER tail. H_1: \mu_0 → reject in BOTH tails. Source: 9709 Examiner Reports 2022-2024

Why is "Concluding $H_0$ is TRUE" a common mistake in Hypothesis tests?

Why it happens: Treating failure-to-reject as proof. How to avoid it: If we don't reject, say 'insufficient evidence to reject H_0', NOT 'H_0 is true'. Source: 9709 Examiner Reports 2022-2024

Cambridge International A Levels Mathematics (9709)

Hypothesis tests

Q: Why is "Concluding $H_0$ is TRUE" a common mistake in Hypothesis tests?

Why it happens: Treating failure-to-reject as proof. How to avoid it: If we don't reject, say 'insufficient evidence to reject H_0', NOT 'H_0 is true'. Source: 9709 Examiner Reports 2022-2024

0% Complete

3 tasks remaining

Short Notes - Hypothesis tests

Detailed Study Notes

Detailed notes on Probability and Statistics 2 - Paper 6 for Cambridge International A Levels Mathematics, covering key concepts, explanations, examples, and exam-focused revision points.

Hypothesis Tests Study Notes — Cambridge International A Level Mathematics 9709 S2 (2024-2026 syllabus)

Formal hypothesis testing. $H_0, H_1$ . Test statistic. Critical values, $p$ -values. Type I and II errors.

At a glance

$H_0$ : null (default).
$H_1$ : alternative.
Reject $H_0$ if test stat in critical region OR $p < \alpha$ .
Type I: reject true $H_0$ . P = $\alpha$ .
Type II: don't reject false $H_0$ . P = $\beta$ .

What you’ll learn

Mapped to the Cambridge International A Level 9709 syllabus (2024-2026).

S2.5.1 — State $H_0$ and $H_1$ .
S2.5.2 — Compute test statistic.
S2.5.3 — Use critical values or $p$ -values.
S2.5.4 — Identify Type I and II errors.

Hypothesis test procedure

Five steps.

Steps.

State hypotheses. $H_0$ (default) and $H_1$ (alternative, one- or two-tailed).
Choose significance level $\alpha$ . Usually 5% or 1%.
Compute test statistic.
Find critical region OR $p$ -value.
Conclude. Reject if in critical region / if $p < \alpha$ . State in context.

One- vs two-tailed.

$H_1: \mu > \mu_0$ : upper tail.
$H_1: \mu < \mu_0$ : lower tail.
$H_1: \mu \ne \mu_0$ : both tails.

Cambridge tip. Always state hypotheses clearly, including in symbolic form.

The significance level is chosen BEFORE the data is seen. The direction of H₁ is what decides whether the critical region is one tail or two — get that wrong and every later step is wasted.

The same 5% is spent differently: all in one tail, or split between two. That is why the two-tailed critical value (1.96) is further out than the one-tailed one (1.645).

Five-step procedure.
Direction of $H_1$ determines tail.
State conclusion in context.

$z$ -test for mean

Known $\sigma$ or large $n$ .

Test statistic. $z = \frac{\bar{x} - \mu_0}{\sigma / \sqrt{n}}$ .

Critical values.

One-tailed 5%: $\pm 1.645$ .
One-tailed 1%: $\pm 2.326$ .
Two-tailed 5%: $\pm 1.96$ .
Two-tailed 1%: $\pm 2.576$ .

Example. $\sigma = 6$ , $n = 36$ , $\bar{x} = 52$ . Test $H_0: \mu = 50$ vs $H_1: \mu > 50$ at 5%.

$z = \frac{52 - 50}{6/6} = 2$ .
Critical value 1.645. $z = 2 > 1.645$ → reject $H_0$ .
Evidence at 5% level that $\mu > 50$ .

Cambridge tip. Show test statistic computation and comparison clearly.

Critical value and p-value are two readings of the same picture and must always agree. Finish by stating the conclusion in context, not just 'reject H₀'.

$z = \frac{\bar{x} - \mu_0}{\sigma / \sqrt{n}}$ .
Compare with critical value.
State conclusion in context.

See the full worked example for hypothesis tests →

Poisson hypothesis tests

$p$ -value from PMF.

Setup. $X \sim \text{Po}(\lambda)$ . Test $H_0: \lambda = \lambda_0$ vs $H_1: \lambda > \lambda_0$ (or $<$ or $\ne$ ).

$p$ -value. Compute from Poisson PMF directly.

Upper-tail: $p = P(X \geq x_{\text{observed}} | H_0)$ .
Lower-tail: $p = P(X \leq x_{\text{observed}} | H_0)$ .

Decision. Reject $H_0$ if $p < \alpha$ .

Example. $\text{Po}(5)$ historically; observed 2 defects. Test $H_0: \lambda = 5$ vs $H_1: \lambda < 5$ at 5%.

$p = P(X \leq 2) = e^{-5}(1 + 5 + 12.5) = e^{-5} \cdot 18.5 \approx 0.1247$ .
$p = 0.1247 > 0.05$ . Do NOT reject $H_0$ . No evidence of improvement.

Cambridge tip. Discrete distributions use exact $p$ -values (no normal approximation needed unless asked).

For a discrete distribution you add the exact probabilities in the tail rather than standardising. Being above α here means two defects is not surprising for a Po(5) process.

Use Poisson PMF directly.
Compare $p$ to $\alpha$ .
Discrete: exact $p$ -value.

See the full worked example for hypothesis tests →

Type I and Type II errors

Two types of mistake.

Type I error. Reject $H_0$ when $H_0$ TRUE.

Probability = $\alpha$ (significance level).
Sometimes called 'false positive'.

Type II error. Don't reject $H_0$ when $H_0$ FALSE.

Probability = $\beta$ .
Sometimes called 'false negative'.

Power = $1 - \beta$ . Probability of correctly rejecting false $H_0$ .

Trade-off. Lower $\alpha$ → higher $\beta$ (less likely to reject anything, including false $H_0$ ). Larger $n$ reduces both.

Cambridge tip. Define Type I and II clearly. State that their PROBABILITIES are $\alpha$ and $\beta$ .

α and β sit on opposite sides of the same critical value, so tightening one loosens the other. Only a bigger sample, which narrows both curves, reduces both errors together.

Type I: reject true.
Type II: don't reject false.
Trade-off between them.

See the full worked example for hypothesis tests →

Quick recap

Five-step procedure.
$z$ -test for mean.
Poisson exact $p$ -values.
Type I/II error definitions.

Memorise this

Verbatim phrases and definitions Cambridge mark schemes credit.

$z$ -test formula.
Critical values for common $\alpha$ .
Type I, II definitions.

How it’s examined

Hypothesis tests are CENTRAL to S2 — typically 10-15 marks. Most-tested: $z$ -test (10 marks), Poisson test (8 marks), Type I/II errors (5 marks).

Sources: Cambridge International A Level Mathematics 9709 syllabus (2024-2026); 9709 Examiner Reports 2022-2024; 9709/62 May/Jun 2024 question paper and mark scheme. Last reviewed 2026-05-11.

Take this whole topic with you

Step-by-step worked examples — Hypothesis tests

Step-by-step solutions to past-paper-style questions on hypothesis tests, written exactly the way a tutor would explain them at the board.

1 $z$ -test for mean (10 marks)
Extended• Adapted from 9709/62 May/Jun 2024• hypothesis test
Question
Population SD $\sigma = 6$ . Sample of $n = 36$ gives $\bar{x} = 52$ . Test $H_0: \mu = 50$ vs $H_1: \mu > 50$ at 5% level. (10 marks)
Step-by-step solution
1. Step 1
  State hypotheses. $H_0: \mu = 50$ ; $H_1: \mu > 50$ .
2. Step 2
  Significance level. $\alpha = 0.05$ . One-tailed. Critical $z = 1.645$ .
3. Step 3
  Test statistic.
  $z = \dfrac{\bar{x} - \mu_0}{\sigma / \sqrt{n}} = \dfrac{52 - 50}{6 / 6} = 2$
4. Step 4
  Compare. $z = 2 > 1.645$ → reject $H_0$ .
5. Step 5
  Conclusion. Evidence at 5% level that $\mu > 50$ .
Answer
$z = 2 > 1.645$ . Reject $H_0$ . Evidence $\mu > 50$ .
2Poisson hypothesis test (8 marks)
Extended• Poisson test
Question
Defects historically follow $\text{Po}(5)$ per unit. After improvements, 1 unit had 2 defects. Test $H_0: \lambda = 5$ vs $H_1: \lambda < 5$ at 5% level. (8 marks)
Step-by-step solution
1. Step 1
  Hypotheses. $H_0: \lambda = 5$ ; $H_1: \lambda < 5$ . One-tailed (lower).
2. Step 2
  $p$ -value. $P(X \leq 2 | \lambda = 5)$ .
3. Step 3
  Compute. $P(X \leq 2) = e^{-5}(1 + 5 + 25/2) = e^{-5} \times 18.5 \approx 0.1247$ .
4. Step 4
  Compare. $p = 0.1247 > 0.05$ → do NOT reject $H_0$ .
5. Step 5
  Conclusion. Insufficient evidence to claim improvement.
Answer
$p \approx 0.1247 > 0.05$ . Do not reject $H_0$ . No significant evidence of improvement.
3Type I and II errors (5 marks)
Extended• errors
Question
Define Type I and Type II error. State their probabilities. (5 marks)
Step-by-step solution
1. Step 1
  Type I error. Rejecting $H_0$ when $H_0$ is TRUE. Probability = $\alpha$ (significance level).
2. Step 2
  Type II error. Failing to reject $H_0$ when $H_0$ is FALSE. Probability = $\beta$ . Depends on actual parameter value.
3. Step 3
  Power = $1 - \beta$ . Probability of correctly rejecting false $H_0$ .
Answer
Type I: reject true $H_0$ , P = $\alpha$ . Type II: don't reject false $H_0$ , P = $\beta$ .

Key Formulae — Hypothesis tests

The formulae you need to memorise for hypothesis tests on the Cambridge International A Level 9709 paper, with every variable defined in plain English and a note on when to use it.

$z$ -test statistic
$z = \dfrac{\bar{x} - \mu_0}{\sigma / \sqrt{n}}$
$\mu_0$
value under $H_0$
When to use
Testing mean with known $\sigma$ or large $n$ .
Critical values
$One-tailed 5%: 1.645; 1%: 2.326. Two-tailed 5%: 1.96; 1%: 2.576$
When to use
Compare with test statistic for rejection region.

Key Definitions and Keywords — Hypothesis tests

Definitions to memorise and the exact keywords mark schemes credit for hypothesis tests answers — sharpened from recent examiner reports for the 2026 Cambridge International A Level 9709 sitting.

Null hypothesis $H_0$
Examiner keyword
Default claim, usually 'no effect' or 'no change'. Tested against alternative.
Alternative hypothesis $H_1$
Examiner keyword
Claim we're trying to support. One- or two-tailed.
Significance level $\alpha$
Examiner keyword
Probability of Type I error. Threshold for rejecting $H_0$ .
$p$ -value
Examiner keyword
Probability of getting test statistic at least as extreme as observed, assuming $H_0$ true. Reject if $p < \alpha$ .
Type I error
Examiner keyword
Rejecting $H_0$ when it is true. Probability = $\alpha$ .
Type II error
Examiner keyword
Failing to reject $H_0$ when it is false. Probability = $\beta$ .

Common Mistakes and Misconceptions — Hypothesis tests

The traps other students keep falling into on hypothesis tests questions — taken from recent Cambridge International A Level 9709 examiner reports and mark schemes — and how to avoid them.

✕Wrong rejection region for one-tailed test
9709 Examiner Reports 2022-2024
Why it happens
Looking at wrong tail.
How to avoid it
$H_1: \mu > \mu_0$ → reject in UPPER tail. $H_1: \mu < \mu_0$ → reject in LOWER tail. $H_1: \mu \ne \mu_0$ → reject in BOTH tails.
✕Concluding $H_0$ is TRUE
9709 Examiner Reports 2022-2024
Why it happens
Treating failure-to-reject as proof.
How to avoid it
If we don't reject, say 'insufficient evidence to reject $H_0$ ', NOT ' $H_0$ is true'.

Hypothesis tests — frequently asked questions

The things students keep getting wrong in this sub-topic, answered.

Cambridge International A Levels Mathematics (9709)

Hypothesis tests

0% Complete

3 tasks remaining

Short Notes - Hypothesis tests

Detailed Study Notes

Detailed notes on Probability and Statistics 2 - Paper 6 for Cambridge International A Levels Mathematics, covering key concepts, explanations, examples, and exam-focused revision points.

Hypothesis Tests Study Notes — Cambridge International A Level Mathematics 9709 S2 (2024-2026 syllabus)

Formal hypothesis testing. $H_0, H_1$ . Test statistic. Critical values, $p$ -values. Type I and II errors.

At a glance

$H_0$ : null (default).
$H_1$ : alternative.
Reject $H_0$ if test stat in critical region OR $p < \alpha$ .
Type I: reject true $H_0$ . P = $\alpha$ .
Type II: don't reject false $H_0$ . P = $\beta$ .

What you’ll learn

Mapped to the Cambridge International A Level 9709 syllabus (2024-2026).

S2.5.1 — State $H_0$ and $H_1$ .
S2.5.2 — Compute test statistic.
S2.5.3 — Use critical values or $p$ -values.
S2.5.4 — Identify Type I and II errors.

Hypothesis test procedure

Five steps.

Steps.

State hypotheses. $H_0$ (default) and $H_1$ (alternative, one- or two-tailed).
Choose significance level $\alpha$ . Usually 5% or 1%.
Compute test statistic.
Find critical region OR $p$ -value.
Conclude. Reject if in critical region / if $p < \alpha$ . State in context.

One- vs two-tailed.

$H_1: \mu > \mu_0$ : upper tail.
$H_1: \mu < \mu_0$ : lower tail.
$H_1: \mu \ne \mu_0$ : both tails.

Cambridge tip. Always state hypotheses clearly, including in symbolic form.

The significance level is chosen BEFORE the data is seen. The direction of H₁ is what decides whether the critical region is one tail or two — get that wrong and every later step is wasted.

The same 5% is spent differently: all in one tail, or split between two. That is why the two-tailed critical value (1.96) is further out than the one-tailed one (1.645).

Five-step procedure.
Direction of $H_1$ determines tail.
State conclusion in context.

$z$ -test for mean

Known $\sigma$ or large $n$ .

Test statistic. $z = \frac{\bar{x} - \mu_0}{\sigma / \sqrt{n}}$ .

Critical values.

One-tailed 5%: $\pm 1.645$ .
One-tailed 1%: $\pm 2.326$ .
Two-tailed 5%: $\pm 1.96$ .
Two-tailed 1%: $\pm 2.576$ .

Example. $\sigma = 6$ , $n = 36$ , $\bar{x} = 52$ . Test $H_0: \mu = 50$ vs $H_1: \mu > 50$ at 5%.

$z = \frac{52 - 50}{6/6} = 2$ .
Critical value 1.645. $z = 2 > 1.645$ → reject $H_0$ .
Evidence at 5% level that $\mu > 50$ .

Cambridge tip. Show test statistic computation and comparison clearly.

Critical value and p-value are two readings of the same picture and must always agree. Finish by stating the conclusion in context, not just 'reject H₀'.

$z = \frac{\bar{x} - \mu_0}{\sigma / \sqrt{n}}$ .
Compare with critical value.
State conclusion in context.

See the full worked example for hypothesis tests →

Poisson hypothesis tests

$p$ -value from PMF.

Setup. $X \sim \text{Po}(\lambda)$ . Test $H_0: \lambda = \lambda_0$ vs $H_1: \lambda > \lambda_0$ (or $<$ or $\ne$ ).

$p$ -value. Compute from Poisson PMF directly.

Upper-tail: $p = P(X \geq x_{\text{observed}} | H_0)$ .
Lower-tail: $p = P(X \leq x_{\text{observed}} | H_0)$ .

Decision. Reject $H_0$ if $p < \alpha$ .

Example. $\text{Po}(5)$ historically; observed 2 defects. Test $H_0: \lambda = 5$ vs $H_1: \lambda < 5$ at 5%.

$p = P(X \leq 2) = e^{-5}(1 + 5 + 12.5) = e^{-5} \cdot 18.5 \approx 0.1247$ .
$p = 0.1247 > 0.05$ . Do NOT reject $H_0$ . No evidence of improvement.

Cambridge tip. Discrete distributions use exact $p$ -values (no normal approximation needed unless asked).

For a discrete distribution you add the exact probabilities in the tail rather than standardising. Being above α here means two defects is not surprising for a Po(5) process.

Use Poisson PMF directly.
Compare $p$ to $\alpha$ .
Discrete: exact $p$ -value.

See the full worked example for hypothesis tests →

Type I and Type II errors

Two types of mistake.

Type I error. Reject $H_0$ when $H_0$ TRUE.

Probability = $\alpha$ (significance level).
Sometimes called 'false positive'.

Type II error. Don't reject $H_0$ when $H_0$ FALSE.

Probability = $\beta$ .
Sometimes called 'false negative'.

Power = $1 - \beta$ . Probability of correctly rejecting false $H_0$ .

Trade-off. Lower $\alpha$ → higher $\beta$ (less likely to reject anything, including false $H_0$ ). Larger $n$ reduces both.

Cambridge tip. Define Type I and II clearly. State that their PROBABILITIES are $\alpha$ and $\beta$ .

α and β sit on opposite sides of the same critical value, so tightening one loosens the other. Only a bigger sample, which narrows both curves, reduces both errors together.

Type I: reject true.
Type II: don't reject false.
Trade-off between them.

See the full worked example for hypothesis tests →

Quick recap

Five-step procedure.
$z$ -test for mean.
Poisson exact $p$ -values.
Type I/II error definitions.

Memorise this

Verbatim phrases and definitions Cambridge mark schemes credit.

$z$ -test formula.
Critical values for common $\alpha$ .
Type I, II definitions.

How it’s examined

Hypothesis tests are CENTRAL to S2 — typically 10-15 marks. Most-tested: $z$ -test (10 marks), Poisson test (8 marks), Type I/II errors (5 marks).

Sources: Cambridge International A Level Mathematics 9709 syllabus (2024-2026); 9709 Examiner Reports 2022-2024; 9709/62 May/Jun 2024 question paper and mark scheme. Last reviewed 2026-05-11.

Take this whole topic with you

Step-by-step worked examples — Hypothesis tests

Step-by-step solutions to past-paper-style questions on hypothesis tests, written exactly the way a tutor would explain them at the board.

1 $z$ -test for mean (10 marks)
Extended• Adapted from 9709/62 May/Jun 2024• hypothesis test
Question
Population SD $\sigma = 6$ . Sample of $n = 36$ gives $\bar{x} = 52$ . Test $H_0: \mu = 50$ vs $H_1: \mu > 50$ at 5% level. (10 marks)
Step-by-step solution
1. Step 1
  State hypotheses. $H_0: \mu = 50$ ; $H_1: \mu > 50$ .
2. Step 2
  Significance level. $\alpha = 0.05$ . One-tailed. Critical $z = 1.645$ .
3. Step 3
  Test statistic.
  $z = \dfrac{\bar{x} - \mu_0}{\sigma / \sqrt{n}} = \dfrac{52 - 50}{6 / 6} = 2$
4. Step 4
  Compare. $z = 2 > 1.645$ → reject $H_0$ .
5. Step 5
  Conclusion. Evidence at 5% level that $\mu > 50$ .
Answer
$z = 2 > 1.645$ . Reject $H_0$ . Evidence $\mu > 50$ .
2Poisson hypothesis test (8 marks)
Extended• Poisson test
Question
Defects historically follow $\text{Po}(5)$ per unit. After improvements, 1 unit had 2 defects. Test $H_0: \lambda = 5$ vs $H_1: \lambda < 5$ at 5% level. (8 marks)
Step-by-step solution
1. Step 1
  Hypotheses. $H_0: \lambda = 5$ ; $H_1: \lambda < 5$ . One-tailed (lower).
2. Step 2
  $p$ -value. $P(X \leq 2 | \lambda = 5)$ .
3. Step 3
  Compute. $P(X \leq 2) = e^{-5}(1 + 5 + 25/2) = e^{-5} \times 18.5 \approx 0.1247$ .
4. Step 4
  Compare. $p = 0.1247 > 0.05$ → do NOT reject $H_0$ .
5. Step 5
  Conclusion. Insufficient evidence to claim improvement.
Answer
$p \approx 0.1247 > 0.05$ . Do not reject $H_0$ . No significant evidence of improvement.
3Type I and II errors (5 marks)
Extended• errors
Question
Define Type I and Type II error. State their probabilities. (5 marks)
Step-by-step solution
1. Step 1
  Type I error. Rejecting $H_0$ when $H_0$ is TRUE. Probability = $\alpha$ (significance level).
2. Step 2
  Type II error. Failing to reject $H_0$ when $H_0$ is FALSE. Probability = $\beta$ . Depends on actual parameter value.
3. Step 3
  Power = $1 - \beta$ . Probability of correctly rejecting false $H_0$ .
Answer
Type I: reject true $H_0$ , P = $\alpha$ . Type II: don't reject false $H_0$ , P = $\beta$ .

Key Formulae — Hypothesis tests

The formulae you need to memorise for hypothesis tests on the Cambridge International A Level 9709 paper, with every variable defined in plain English and a note on when to use it.

$z$ -test statistic
$z = \dfrac{\bar{x} - \mu_0}{\sigma / \sqrt{n}}$
$\mu_0$
value under $H_0$
When to use
Testing mean with known $\sigma$ or large $n$ .
Critical values
$One-tailed 5%: 1.645; 1%: 2.326. Two-tailed 5%: 1.96; 1%: 2.576$
When to use
Compare with test statistic for rejection region.

Key Definitions and Keywords — Hypothesis tests

Definitions to memorise and the exact keywords mark schemes credit for hypothesis tests answers — sharpened from recent examiner reports for the 2026 Cambridge International A Level 9709 sitting.

Null hypothesis $H_0$
Examiner keyword
Default claim, usually 'no effect' or 'no change'. Tested against alternative.
Alternative hypothesis $H_1$
Examiner keyword
Claim we're trying to support. One- or two-tailed.
Significance level $\alpha$
Examiner keyword
Probability of Type I error. Threshold for rejecting $H_0$ .
$p$ -value
Examiner keyword
Probability of getting test statistic at least as extreme as observed, assuming $H_0$ true. Reject if $p < \alpha$ .
Type I error
Examiner keyword
Rejecting $H_0$ when it is true. Probability = $\alpha$ .
Type II error
Examiner keyword
Failing to reject $H_0$ when it is false. Probability = $\beta$ .

Common Mistakes and Misconceptions — Hypothesis tests

The traps other students keep falling into on hypothesis tests questions — taken from recent Cambridge International A Level 9709 examiner reports and mark schemes — and how to avoid them.

✕Wrong rejection region for one-tailed test
9709 Examiner Reports 2022-2024
Why it happens
Looking at wrong tail.
How to avoid it
$H_1: \mu > \mu_0$ → reject in UPPER tail. $H_1: \mu < \mu_0$ → reject in LOWER tail. $H_1: \mu \ne \mu_0$ → reject in BOTH tails.
✕Concluding $H_0$ is TRUE
9709 Examiner Reports 2022-2024
Why it happens
Treating failure-to-reject as proof.
How to avoid it
If we don't reject, say 'insufficient evidence to reject $H_0$ ', NOT ' $H_0$ is true'.

Hypothesis tests — frequently asked questions

The things students keep getting wrong in this sub-topic, answered.

Hypothesis tests

Short Notes - Hypothesis tests

Detailed Study Notes

At a glance

What you’ll learn

Hypothesis test procedure

zzz-test for mean

Poisson hypothesis tests

Type I and Type II errors

Quick recap

Memorise this

How it’s examined

Take this whole topic with you

1zzz-test for mean (10 marks)

2Poisson hypothesis test (8 marks)

3Type I and II errors (5 marks)

zzz-test statistic

Critical values

Null hypothesis H0H_0H0​

Alternative hypothesis H1H_1H1​

Significance level α\alphaα

ppp-value

Type I error

Type II error

✕Wrong rejection region for one-tailed test

✕Concluding H0H_0H0​ is TRUE

Hypothesis tests — frequently asked questions

Hypothesis tests

Short Notes - Hypothesis tests

Detailed Study Notes

At a glance

What you’ll learn

Hypothesis test procedure

zzz-test for mean

Poisson hypothesis tests

Type I and Type II errors

Quick recap

Memorise this

How it’s examined

Take this whole topic with you

1zzz-test for mean (10 marks)

2Poisson hypothesis test (8 marks)

3Type I and II errors (5 marks)

zzz-test statistic

Critical values

Null hypothesis H0H_0H0​

Alternative hypothesis H1H_1H1​

Significance level α\alphaα

ppp-value

Type I error

Type II error

✕Wrong rejection region for one-tailed test

✕Concluding H0H_0H0​ is TRUE

Hypothesis tests — frequently asked questions

$z$ -test for mean

1 $z$ -test for mean (10 marks)

$z$ -test statistic

Null hypothesis $H_0$

Alternative hypothesis $H_1$

Significance level $\alpha$

$p$ -value

✕Concluding $H_0$ is TRUE

$z$ -test for mean

1 $z$ -test for mean (10 marks)

$z$ -test statistic

Null hypothesis $H_0$

Alternative hypothesis $H_1$

Significance level $\alpha$

$p$ -value

✕Concluding $H_0$ is TRUE