Did you know? A p-value of 0.0228 does not mean there is a 97.72% chance that your claim is correct. That interpretation sounds persuasive—and it is wrong.
In JC2 H1 Mathematics, hypothesis testing is not only about getting a number from the graphing calculator. It is about deciding what the evidence permits you to say, then writing the conclusion in the context of the question.
The meaning of a p-value in one sentence
A p-value tells you how likely a result at least as extreme as the one observed would be if the null hypothesis were true, under the test’s assumptions.
It does not directly measure the probability that the null hypothesis is true. And it does not prove the alternative hypothesis beyond doubt.
A worked H1 example in five steps
Suppose a service claims its population mean waiting time is 50 minutes. A random sample of 64 waits has mean 52 minutes. Assume the population standard deviation is known to be 8 minutes and the conditions for the test are met. We want to test whether the mean has increased.
- 1. State hypotheses: H₀: μ = 50; H₁: μ > 50.
- 2. Identify the tail: the phrase ‘has increased’ makes this a one-tailed upper test.
- 3. Calculate the statistic: standard error = 8/√64 = 1, so z = (52 − 50)/1 = 2.
- 4. Find the p-value: P(Z ≥ 2) ≈ 0.0228 for Z ~ N(0, 1).
- 5. Decide and conclude: 0.0228 < 0.05, so reject H₀ at the 5% significance level.
A strong final line would read: ‘At the 5% significance level, there is sufficient evidence that the population mean waiting time exceeds 50 minutes.’
Notice the key words: population mean, exceeds, and 50 minutes. They connect the statistical decision back to what the question actually asked.
What changes if the significance level is 1%?
The same p-value, 0.0228, is now greater than 0.01. We therefore fail to reject H₀ at the 1% significance level. We do not say that the null hypothesis has been proven true.
The sample data did not change. The decision threshold changed. This is why every written conclusion should name the significance level.
A 20-second checklist before moving on
- Is the alternative hypothesis one-sided or two-sided?
- Did I use the sampling distribution and standard error rather than the raw observation distribution?
- Did I compare p with the given significance level, not an imagined default?
- Did I write ‘fail to reject’ rather than ‘accept’ when evidence is insufficient?
- Does my final sentence answer the real-world claim?
The SEAB H1 Mathematics 8865 syllabus includes hypothesis tests for a population mean, one-tail and two-tail tests, and interpreting results in context. Those are the skills worth practising, not just memorising a calculator sequence.
Quick questions
Does a very small p-value show a large practical effect? Not necessarily. Evidence against a null claim and the size or importance of a difference are separate questions.
What can a JC2 H1 Mathematics tutor do here? Look at whether the student is losing marks on model choice, tail direction, calculator input or the final sentence—then repair that exact step.
Read the full H1 Mathematics topic guide, and ask eduKate about JC2 H1 support if your recent papers show repeated hypothesis-testing errors.