IB Mathematics AI AHL A linear combination of n independent normal MAI Study Notes - New Syllabus
IB Mathematics AI AHL A linear combination of n independent normal MAI Study Notes
LEARNING OBJECTIVE
- Central limit theorem.
Key Concepts:
- Central limit theorem.
- A linear combination of n independent normal random variables
- IBDP Maths AI SL- IB Style Practice Questions with Answer-Topic Wise-Paper 1
- IBDP Maths AI SL- IB Style Practice Questions with Answer-Topic Wise-Paper 2
- IB DP Maths AI HL- IB Style Practice Questions with Answer-Topic Wise-Paper 1
- IB DP Maths AI HL- IB Style Practice Questions with Answer-Topic Wise-Paper 2
- IB DP Maths AI HL- IB Style Practice Questions with Answer-Topic Wise-Paper 3
THE CENTRAL LIMIT THEOREM
The Central Limit Theorem (CLT)
Let $X_1, X_2, \ldots, X_n$ be i.i.d. random variables with expected value $\mathbb{E}X_i = \mu < \infty$ and variance $0 < \text{Var}(X_i) = \sigma^2 < \infty$. Then, the random variable
$Z_n = \frac{\overline{X} – \mu}{\sigma/\sqrt{n}} = \frac{X_1 + X_2 + \ldots + X_n – n\mu}{\sqrt{n}\sigma}$
converges in distribution to the standard normal random variable as $n$ goes to infinity, that is
$\lim_{n \to \infty} P(Z_n \leq x) = \Phi(x), \qquad \text{for all } x \in \mathbb{R},$
where $\Phi(x)$ is the standard normal CDF.
Let $X$ be a random variable with mean $\mu$ and standard deviation $\sigma$.
The sample mean $\overline{X}$ for $n$ independent samples has:
$E(\overline{X}) = \mu, \quad \text{Var}(\overline{X}) = \frac{\sigma^2}{n}, \quad \text{SD}(\overline{X}) = \frac{\sigma}{\sqrt{n}}$
Linear Combinations of Normal Random Variables
If $X \sim N(\mu, \sigma^2)$, then any linear combination, such as $\overline{X}$, is also Normally distributed:
$\overline{X} \sim N\left( \mu, \frac{\sigma^2}{n} \right)$
Example Suppose $X \sim N(70, 9)$. A sample of $n = 16$ gives $\overline{X} \sim N\left(70, \frac{9}{16} \right) = N(70, 0.5625)$. Find $P(69 < \overline{X} < 71)$. ▶️Answer/ExplanationSolution: Using a calculator: $P(69 < \overline{X} < 71) = 0.954$ |
Statement (CLT)
For large sample sizes ($n > 30$), regardless of the original distribution of $X$, the distribution of the sample mean $\overline{X}$ approaches a Normal distribution:
$
\overline{X} \sim N\left( \mu, \frac{\sigma^2}{n} \right)
$
Example Let $X \sim \text{Poisson}(10)$, $n = 40$. Then: $ Find $P(9 < \overline{X} < 11)$: ▶️Answer/ExplanationSolution: $P(9 < \overline{X} < 11) = 0.954$ The approximation improves as $n$ increases. $n > 30$ is typically considered sufficient. |
Example Let $X \sim \text{Binomial}(100, 0.2)$, $n = 40$. Then $\mu = 20$, $\sigma^2 = 16$, so: $ Find $P(\overline{X} < 19) = ?$ ▶️Answer/ExplanationSolution: $P(\overline{X} < 19) = 0.0569$ |
Using the Z-Table
Convert to a standard normal variable:
$
Z = \frac{\overline{X} – \mu}{\sigma / \sqrt{n}}
$
Example Population: $\mu = 100$, $\sigma = 30$, $n = 50$ $\overline{X} \sim N(100, 18)$ Find $P(90 < \overline{X} < 110)$: ▶️Answer/ExplanationSolution: Convert to $Z$-scores and use Z-table or calculator: $P = 0.982$ |
Applications
- Estimating population parameters
- Hypothesis testing
- Confidence intervals
- Valid for skewed/non-normal distributions when $n$ is large
Example You want to test whether a new teaching method increases test scores. The current average is 70 with $\sigma = 12$. A sample of 40 students has a mean of 74. ▶️Answer/ExplanationSolution: $ Significant evidence that the method improves scores. |
Visualizing Central Limit Theorem & Practical Considerations
- Histograms of sample means become bell-shaped as $n$ increases
- Demonstrated using simulation
- Independence: samples must be independent
- Finite population correction if sampling without replacement
- Not valid for very small samples from non-Normal populations
Example (conceptual):
Simulate 10,000 samples of size $n = 5$, $n = 30$, and $n = 100$ from a skewed distribution.
Plot histograms of sample means.
→ Distribution becomes approximately normal as $n$ increases.
Example:
For a small sample $n = 10$ from a skewed population, the CLT may not apply. Use non-parametric methods or increase $n$.