QCE Mathematical Methods - Unit 4 - Continuous random variables and the normal distribution

Continuous Random Variables | QCE Mathematical Methods

Understand QCE Mathematical Methods continuous random variables, density functions, cumulative distribution functions, expected value and variance.

Updated 2026-05-18 - 4 min read

QCAA official coverage - Mathematical Methods 2025 v1.3

Exact syllabus points covered

  1. Use relative frequencies and histograms obtained from data to estimate probabilities associated with a continuous random variable.
  2. Understand probability density functions, cumulative distribution functions, and probabilities associated with a continuous random variable given by integrals.
  3. Calculate the expected value, $E(X)=\mu=\int_{-\infty}^{\infty}xp(x)\,dx$, of a continuous random variable where $p(x)$ is the probability density function.
  4. Calculate the variance, $\operatorname{Var}(X)=\sigma^2=\int_{-\infty}^{\infty}(x-\mu)^2p(x)\,dx$, and standard deviation $\sigma$, of a continuous random variable.
  5. Understand standardised normal variables ($z$-values, $z$-scores) and use these to compare samples.

A continuous random variable can take any value in an interval. Time, height, mass, distance and temperature are common examples. Unlike a discrete random variable, a continuous variable is not handled by adding probabilities at individual points.

For a continuous random variable:

$ P(X=a)=0 $

That does not mean values cannot occur. It means probability is measured across intervals, not exact points.

Continuous density and CDF

Original Sylligence diagram for continuous density cdf.

Continuous density and CDF

Probability density functions

A probability density function, usually written $p(x)$ or $f(x)$, gives a curve whose area represents probability. The total area under the curve must be:

$ \int_{-\infty}^{\infty}p(x)\,dx=1 $

For an interval:

$ P(a\le X\le b)=\int_a^b p(x)\,dx $

A function is a valid probability density function when:

  • $p(x)\ge 0$ over its domain
  • the total area under the curve is $1$
  • probabilities are found by integrating over intervals

If the density is defined only on a finite interval, it is treated as $0$ outside that interval. For example, if $p(x)=2x$ for $0\le x\le 1$, then $p(x)=0$ for $x<0$ and $x>1$.

From histograms to density

Real continuous data often starts as a relative-frequency histogram. In a density histogram, each bar area represents relative frequency:

$ \text{bar area}=\text{relative frequency} $

Because:

$ \text{bar area}=\text{width}\times\text{height} $

the density height is:

$ \text{density}=\frac{\text{relative frequency}}{\text{class width}} $

This is why density can be greater than $1$ for a narrow interval. The height is not the probability; the area is.

Cumulative distribution functions

The cumulative distribution function, $F(x)$, gives the probability that the random variable is less than or equal to a value:

$ F(x)=P(X\le x) $

In integral form:

$ F(x)=\int_{-\infty}^{x}p(t)\,dt $

For a continuous distribution:

$ P(a\le X\le b)=F(b)-F(a) $

This is the statistics version of the definite-integral idea: accumulated probability up to $b$ minus accumulated probability up to $a$.

Because $P(X=a)=0$, the endpoint style does not change the answer:

$ P(a<X<b)=P(a\le X\le b)=P(a<X\le b) $

That is different from discrete random variables, where including or excluding an endpoint can matter.

Expected value and variance

For a continuous random variable:

$ E(X)=\mu=\int_{-\infty}^{\infty}xp(x)\,dx $

Variance is:

$ \operatorname{Var}(X)=\sigma^2=\int_{-\infty}^{\infty}(x-\mu)^2p(x)\,dx $

Standard deviation is $\sigma$.

The shortcut formula also works:

$ \operatorname{Var}(X)=E(X^2)-[E(X)]^2 $

where:

$ E(X^2)=\int_{-\infty}^{\infty}x^2p(x)\,dx $

For functions of a continuous random variable, the expected value is found by integrating the function times the density:

$ E(g(X))=\int_{-\infty}^{\infty}g(x)p(x)\,dx $

Worked example

Worked validation example

Histograms and relative frequency

Histograms from real data can estimate the shape of a continuous distribution. A large sample with sensible class widths can make the distribution easier to see. A small sample or poor bin choice can hide the shape.

For interval questions involving infinity, use the same area idea:

$ P(X>a)=1-F(a) $

and:

$ P(X\le a)=F(a) $

If a density is given in pieces, integrate over the piece or pieces that overlap the interval. Sketching the support first helps you avoid integrating over a region where the density is actually $0$.

Common mistake

Quick check

Sources