Statistical Hypothesis Testing — OCR A-Level Mathematics
Statistical Hypothesis Testing explained
Before a sample is taken, the value of the sample mean is unknown, so X̄ is a random variable: different random samples of the same size n give different values of X̄.
Read the full explanation
Once the sample is drawn, the observed value x̄ is a single realised number, no longer random. Because X̄ is a random variable it has a distribution, the sampling distribution, with mean E(X̄) = μ and variance Var(X̄) = σ²/n, so the standard error is σ/√n. For example, samples of size n = 25 from a population with σ = 10 give a standard error of 10/5 = 2. This distribution underpins the calculation of probabilities and critical regions in a hypothesis test.
Your focus
- Explain why the sample mean is a random variable before the sample is observed.
- State the mean and variance of the sampling distribution of X̄.
- Calculate and interpret the standard error σ/√n in context.
Statistical Hypothesis Testing exam tips
Marking Points
- States that X̄ varies from sample to sample, so it is a random variable before the sample is observed.
- Distinguishes the random variable X̄ from the realised value x̄ obtained from one particular sample.
- Gives the sampling distribution of X̄ with mean μ and variance σ²/n.
- Uses the standard error σ/√n to compute probabilities or critical values for X̄.
Examiner Tips
- 💡Write E(X̄) = μ and Var(X̄) = σ²/n explicitly before substituting numbers.
- 💡State the sample size n when calculating the standard error, since it controls the spread of X̄.
- 💡Keep the distinction between the random variable X̄ and its observed value clear in written answers.
Common Mistakes
- Treats the observed sample mean as fixed and non-random; recognise that X̄ is random before sampling and only the realised value is fixed.
- Uses the population variance σ² instead of σ²/n for the variance of X̄; divide by the sample size to obtain the variance of the mean.
- Forgets to take the square root when finding the standard error; compute σ/√n rather than σ²/n.