The Ultimate Guide to Binomial and Sampling Distributions in HSC Maths Ext 1

Introduction

Have you ever wondered how pollsters predict election results from a small sample of voters? Or how factories calculate the exact probability of producing a defective part? These real-world problems are solved using the mathematics of distributions.

In the Year 12 Mathematics Extension 1 course, the "Binomial distribution and sampling distribution of the mean" topic elevates your understanding of probability from simple dice rolls to complex statistical models. This guide will walk you through the journey from a single Bernoulli trial to the profound implications of the Central Limit Theorem.

Executive Summary

This final statistics module (aligned with outcome ME1-12-06) introduces students to discrete probability distributions (Bernoulli and Binomial) and continuous sampling distributions. You will learn to calculate expected values (means) and variances for these models. More importantly, you will discover the Central Limit Theorem, which states that regardless of the original population's shape, the distribution of sample means will always approximate a normal distribution if the sample size is large enough.

What is this about?

This detailed guide is broken down into three key statistical areas:

Bernoulli Distributions: The foundation block—modelling a single trial with only two possible outcomes (success or failure).
Binomial Distributions: Expanding the model to a fixed number of independent Bernoulli trials and calculating probabilities using combinatorics ( $^nC_r$ ).
Sampling Distribution of the Mean & the Central Limit Theorem (CLT): Understanding how the means of random samples behave and using the normal distribution to estimate probabilities for these sample means.

Let's explore the world of statistical distributions.

Main Content

1. Bernoulli Distributions

A Bernoulli trial is the simplest statistical experiment. It has exactly two possible outcomes: "Success" (which we assign the value $X=1$ ) and "Failure" (assigned the value $X=0$ ).

The probability of success is $p$ .
The probability of failure is $q$ , where $q = 1 - p$ .

The Bernoulli distribution is simply the probability distribution of this single trial.

Expected Value (Mean): $E(X) = \mu = p$
Variance: $Var(X) = p(1 - p) = pq$

Example: Flipping a biased coin where the probability of landing on Heads (Success) is 0.6. The mean is 0.6, and the variance is $0.6 \times 0.4 = 0.24$ .

2. Binomial Distributions

If we take our single Bernoulli trial and repeat it $n$ times (where the trials are independent and the probability $p$ remains constant), we create a Binomial Experiment. The binomial random variable, $X$ , represents the total number of successes in those $n$ trials. We write this as $X \sim Bin(n, p)$ .

Calculating Probabilities: To find the exact probability of getting exactly $r$ successes out of $n$ trials, we use the formula derived from the binomial expansion: $P(X = r) = {}^nC_r \, p^r \, (1 - p)^{n - r}$

Here, $^nC_r$ calculates the number of different ways those $r$ successes can be arranged among the $n$ trials.

Mean and Variance: Because a binomial distribution is just $n$ independent Bernoulli trials added together, the mean and variance are simply multiplied by $n$ :

Expected Value (Mean): $E(X) = \mu = np$
Variance: $Var(X) = np(1 - p) = npq$

Context check: Drawing cards with replacement is a binomial experiment (the probability remains constant). Drawing cards without replacement is NOT binomial (the probability changes after each draw).

3. Sampling Distribution of the Mean and the Central Limit Theorem

In the real world, we rarely know the true mean of an entire population ( $\mu$ ). Instead, we take a random sample of size $n$ , calculate its mean (the sample mean, denoted as $\bar{X}$ ), and use it to estimate the population mean.

If you were to take every possible sample of size $n$ and plot all their means, you would create the Sampling Distribution of the Mean.

The Central Limit Theorem (CLT): This is arguably one of the most important theorems in all of statistics. The CLT states that provided your sample size is large enough (typically $n \ge 30$ ), the sampling distribution of the mean will be approximately normally distributed, regardless of what the original population looked like.

If the original population has a mean $\mu$ and variance $\sigma^2$ , the sampling distribution $\bar{X}$ will have:

Mean: $E(\bar{X}) = \mu$
Variance: $Var(\bar{X}) = \frac{\sigma^2}{n}$

We write this mathematically as: $\bar{X} \approx N(\mu, \frac{\sigma^2}{n})$ .

Because the variance of the sample mean is divided by $n$ , as your sample size gets bigger, the variance gets smaller. This proves mathematically that larger samples provide more accurate, tightly-clustered estimates of the true population mean.

Once you establish that $\bar{X}$ is normally distributed via the CLT, you can use standard z-scores (from the Advanced course) to calculate probabilities relating to the sample mean.

mini-FAQ page

Q: Why do we use $n \ge 30$ for the Central Limit Theorem? A: It is a general statistical rule of thumb. For heavily skewed populations, you might need a larger $n$ for the sampling distribution to look perfectly normal. However, for HSC purposes, if a question states $n \ge 30$ , you are safe to assume the normal approximation holds.

Q: Do I need to use the normal approximation to the binomial distribution? A: No! The current syllabus explicitly excludes the normal approximation to the binomial distribution. If you are dealing with a binomial distribution, you should calculate probabilities directly using the binomial formula or a calculator. You only use the normal distribution when dealing with the sampling distribution of the mean (CLT).

Q: How is the sample variance different from the population variance? A: Be very careful with notation. The population variance is $\sigma^2$ . The variance of the sampling distribution of the mean is $\frac{\sigma^2}{n}$ . Make sure you divide by $n$ when calculating z-scores for sample means!

Common mistakes to avoid

Forgetting to divide variance by $n$ in CLT questions: When calculating a z-score for a sample mean, the standard deviation you must use is $\frac{\sigma}{\sqrt{n}}$ . Using just $\sigma$ is the most common error in this topic.
Confusing binomial scenarios with non-binomial ones: Always check if the trials are independent. If an experiment is "without replacement," the probability changes each trial, meaning it fails the strict definition of a binomial distribution.
Miscalculating $1-p$ : It sounds simple, but under exam pressure, students often miscalculate the failure probability $q$ . Double-check your basic arithmetic!

Practice on Vu's Maths Hub

Probability and statistics are topics where reading the question carefully is half the battle.

Sharpen your statistical intuition with our expert resources on Vu's Maths Hub:

Practice identifying and calculating Binomial probabilities and CLT z-scores with our HSC Distributions booklet.
Build your foundational combinatorics skills required for $^nC_r$ with the HSC Probability booklet.
Check out our Worked Solutions to see how to lay out your statistical reasoning clearly for the examiner.

Connect with me

Ready to conquer distributions and statistics? Join Vu's Maths Hub today and gain access to our extensive collection of Maths Booklets, Worked Solutions, and Trial Papers tailored specifically to help you dominate the NSW HSC curriculum.