Hypothesis Testing Explained: Making Binomial Test Decisions in Exams

Hypothesis Testing Explained: Writing a Clear, Examiner-Approved Conclusion

📊 Why Hypothesis Testing Loses Marks So Easily

Hypothesis testing looks like a checklist topic. That’s the trap. Students learn a “process”, then panic when the wording changes and the process stops feeling automatic.

What examiners actually mark is the logic. They want to see a claim turned into hypotheses, a probability model selected, and a decision made using a stated significance level. If any link in that chain is weak, marks fall away fast. This is one of those areas where A Level Maths made clearer is not about extra practice questions, but about cleaner structure.

One more thing: examiners hate vague conclusions. If the conclusion doesn’t match the hypotheses, it doesn’t score well.

🔙 Previous topic:

Before making decisions in hypothesis testing, students need confidence in statistical distributions, since choosing the correct model underpins every binomial test.

🧠 What a Hypothesis Test Really Is

A hypothesis test is a method for deciding whether sample evidence is strong enough to doubt a claim about a population.

It does not prove anything. It does not “show” the claim is true or false. It checks whether the sample result would be unlikely if the claim were true. That’s it. The word “unlikely” is where the significance level comes in.

At A Level Statistics, the most common hypothesis tests are built around binomial models (success/failure). This blog keeps it there, on purpose.

🧾 Writing H₀ and H₁ Properly

✅ The null hypothesis (H₀)

The null hypothesis is the default claim. It must contain an equality.

H₀: p = p₀

✅ The alternative hypothesis (H₁)

The alternative hypothesis is what you are looking for evidence of. It must match the wording of the question.

Greater than
H₁: p > p₀

Less than
H₁: p < p₀

Different from
H₁: p ≠ p₀

A lot of marks are lost through poorly written hypotheses. Examiners cannot award method marks if the hypotheses are unclear or incorrectly stated.

🎯 Significance Level: 5% or 10%

In most A Level questions, the significance level is given. If it is not stated, the standard assumption is 5%.

α = 0.05

This means we treat outcomes with probability less than 0.05 under H₀ as sufficiently unlikely.

It does not mean there is a 5% chance that H₀ is true. That phrasing is penalised in marking.

🔢 The Binomial Model (The Core Maths)

When you have:

a fixed number of trials (n)
two outcomes (success or failure)
a constant probability (p)
independent trials

then, under H₀, the test statistic is:

X ~ Bin(n, p₀)

This is the mathematical heart of many A Level hypothesis tests. Get this line right and the question becomes much calmer.

🧪 Worked Example: Binomial Hypothesis Test (Full Solution)

A basketball player claims they score a free throw with probability (0.5). In 20 attempts, they score 15.
Test at the 5% significance level whether the player’s success probability is greater than 0.5.

🧾 Step 1: Define hypotheses

H₀: p = 0.5
H₁: p > 0.5

This is a one-tailed test because we are only checking “greater than”.

🔢 Step 2: Define the distribution under H₀

Let X be the number of successful free throws in 20 attempts.

X ~ Bin(20, 0.5)

Observed value: x = 15

🎯 Step 3: Use a 5% critical region (right tail)

Because H₁: p > 0.5, we look for large values of X.

We need the smallest value k such that:
P(X ≥ k) ≤ 0.05

Check:
P(X ≥ 15) = 0.0207

So a valid 5% critical region is:
X ≥ 15

This is the point where many students waffle. Examiners want the critical region stated clearly.

✅ Step 4: Make the decision

We observed x = 15, which lies in the critical region X ≥ 15.
So we reject H₀ at the 5% significance level.

Other Related Topics

Some hypothesis test questions are based on large data sets and require careful interpretation.

The final step in a hypothesis test is writing a clear conclusion based on the result.

Once the structure of $H_0$ and $H_1$ is introduced, it is important to understand where students typically go wrong. The most frequent wording and notation errors are examined in Hypothesis Testing: Common Errors When Stating Hypotheses.

After calculating a test statistic, the next step is making a correct comparison with the critical region. This decision-making process is broken down clearly in Hypothesis Testing Exam Technique: Comparing Test Statistics to Critical Values.

Even when calculations are correct, marks are often lost in the final written conclusion. The structure examiners expect is explained in Hypothesis Testing: Why Students Lose Marks in the Final Conclusion.

Selecting the appropriate distribution is the foundation of a valid hypothesis test. Guidance on making that choice appears in Hypothesis Testing Exam Technique: Choosing the Correct Distribution.

🧠 Step 5: Write a conclusion in context

Reject H₀. There is sufficient evidence at the 5% level to suggest the player’s success probability is greater than 0.5.

That final sentence matters. Examiners do not want “Reject H₀” on its own. They want the claim translated back into English.

📝 How Examiners Really Award Marks (Binomial Test)

Here is how this typically breaks down in mark schemes:

M1 for correct hypotheses, including the correct inequality and correct parameter p.
If the hypotheses are wrong or vague, method marks are limited from the start.

M1 for stating the correct binomial model under H₀:
X ~ Bin(n, p₀)
This is a key method step.

A1 for a correct probability calculation or a correct critical region.
This can be done by:
– calculating P(X ≥ x) directly, or
– finding and stating a correct critical region.

A1 for a correct decision and conclusion:
– reject or do not reject H₀ correctly, and
– include a context sentence that matches H₁.

Examiners are much more generous when they can see a clear structure.
They are much harsher when the conclusion does not match the hypotheses.

🔧 Where Revision Often Goes Wrong

Students practise the button-pressing side of binomial probabilities, then forget the decision side. The decision is where marks are gained and lost.

This is why A Level Maths revision shortcut methods sometimes backfire: students learn quick probability steps but don’t learn to write a clean critical region or a clean conclusion.

A short, disciplined structure beats speed every time in hypothesis testing.

⚠️ Common Errors Examiners See Repeatedly

Students often:

write H₀: p > 0.5, which is wrong because the null hypothesis must contain an equality
use the wrong tail, for example a right-tailed test when a left-tailed test is required
compare the probability to 0.5 instead of the 5% significance level, 0.05
write “accept H₀” instead of the preferred wording “do not reject H₀”
give a conclusion that does not match the alternative hypothesis

None of these mistakes are difficult. They come from rushing the structure rather than misunderstanding the mathematics.

✏️Author Bio

S. Mahandru is an experienced A Level Maths teacher and approved examiner-style tutor with over 15 years’ experience, specialising in clear exam structure, probability decisions, and mark scheme interpretation.

🧭 Next topic:

After making decisions using binomial tests, it’s useful to focus on interpreting data visually, which is developed next in cumulative frequency graphs.

🎯 Final Thought

Hypothesis testing becomes reliable when it is treated as structure rather than a trick. Write the hypotheses clearly, set up the binomial model under H₀, state a 5% critical region, and then draw a careful conclusion. That discipline is exactly what an A Level Maths revision course with worked examples is designed to develop across Statistics.

❓ FAQs — Hypothesis Testing Explained

🧮 Why must H₀ contain an equality?

The null hypothesis must contain an equality because it provides a fixed reference point for the probability model. Hypothesis testing works by assuming H₀ is true and then assessing how unusual the observed result would be under that assumption. Without an equality, there is no single value to test against, so the distribution under H₀ is not properly defined. This is not a matter of wording or style; it is a mathematical requirement. Examiners check this very early in the marking process, often before looking at any calculations. If H₀ is written incorrectly, the number of marks available is usually restricted from the start. Even accurate probability calculations cannot fully compensate for a poorly defined null hypothesis.

📝 Do I ever write “accept H₀”?

Examiners generally prefer the phrase “do not reject H₀” rather than “accept H₀”. Hypothesis testing does not prove that the null hypothesis is true; it only assesses whether there is sufficient evidence against it. The word “accept” suggests a level of certainty that the method does not justify. In practice, this is more about examiner expectation than strict mathematical language. Some mark schemes will tolerate “accept”, but it becomes risky if other parts of the conclusion are weak. Using cautious wording makes it easier for examiners to award the final accuracy mark. When under exam pressure, “do not reject H₀” is the safer and more reliable choice.

🎯 How do I know if the critical region is in the left or right tail?

The position of the critical region is determined entirely by the alternative hypothesis. If H₁ uses the wording “greater than”, the critical region lies in the right tail. If it uses “less than”, the critical region lies in the left tail. If the alternative hypothesis uses “not equal to”, the test is two-tailed and the significance level must be split between both tails. This decision must be made before any probabilities are calculated. It is a decision point rather than a calculation step, and examiners mark it accordingly. Choosing the tail after seeing the sample result is a common error and is penalised consistently. Taking a moment to decide the tail correctly often saves marks later in the question.