Edexcel A-Level Further Mathematics Further Statistics: discrete distributions, Poisson, chi-squared and waiting times

What Further Statistics demands

The Further Statistics options extend statistical work beyond A-Level Mathematics and are taken as one or both of the two optional 9FM0 papers. They reward clear modelling and disciplined hypothesis testing: you must choose the right distribution, compute its mean and variance, and lay out a test with explicit hypotheses, statistic, critical value and conclusion in context. This guide walks through the four headline topics and the exam patterns Edexcel repeats. Each topic has a matching dot-point page with worked questions.

Discrete probability distributions

Discrete probability distributions define a random variable through its probability distribution and compute expectation $E(X) = \sum x P(X = x)$ and variance $\operatorname{Var}(X) = E(X^2) - [E(X)]^2$. Linear coding gives $E(aX + b) = aE(X) + b$ and $\operatorname{Var}(aX + b) = a^2\operatorname{Var}(X)$, with the constant shifting the mean but not the spread.

Poisson and binomial

Poisson and binomial uses the Poisson distribution $\operatorname{Po}(\lambda)$ with $P(X = x) = e^{-\lambda}\frac{\lambda^x}{x!}$ and mean and variance both $\lambda$, the binomial with mean $np$ and variance $np(1 - p)$, the additive property of independent Poisson variables, and the Poisson approximation to the binomial when $n$ is large and $p$ small.

Chi-squared tests

Chi-squared tests compare observed and expected frequencies with the statistic $\chi^2 = \sum \frac{(O - E)^2}{E}$, for goodness of fit (degrees of freedom reduced by one for each estimated parameter) and for independence in contingency tables (degrees of freedom $(r - 1)(c - 1)$), merging classes so each expected frequency is at least $5$ and applying Yates' correction for one degree of freedom.

Geometric and negative binomial

Geometric and negative binomial model waiting times: the geometric counts trials to the first success with mean $\frac{1}{p}$, and the negative binomial counts trials to the $r$th success with mean $\frac{r}{p}$, the geometric being the special case $r = 1$.

How the Further Statistics papers are examined

A typical Edexcel profile:

• Short calculation questions. A Poisson probability, a binomial mean and variance, or a geometric expectation.
• Full hypothesis tests. Chi-squared goodness of fit or a contingency table, written with hypotheses, statistic, critical value and conclusion.
• Modelling and approximation. Choosing a distribution to model a situation, and justifying the Poisson approximation.
• Reasoning in context. Interpreting results for the real scenario, not just stating a number.

Check your knowledge

Attempt these under timed conditions, then check the solutions.

1. For a discrete $X$ with $E(X) = 5$, find $E(3X - 2)$. (1 mark)
2. For the same $X$ with $\operatorname{Var}(X) = 2$, find $\operatorname{Var}(3X - 2)$. (2 marks)
3. State the mean and variance of $\operatorname{Po}(\lambda)$. (2 marks)
4. State the mean and variance of $B(n, p)$. (2 marks)
5. Write the chi-squared test statistic. (1 mark)
6. State the degrees of freedom for a $2 \times 3$ contingency table. (1 mark)
7. State the mean of $\operatorname{Geo}(p)$. (1 mark)
8. State the mean of the negative binomial for the $r$th success. (1 mark)