Normal Distribution And Probability Computation

1. Core Concept Overview & Intuition

The normal distribution is the central continuous probability model of statistics. It describes a random variable whose probability mass is concentrated symmetrically about a central value, with frequencies decaying smoothly and rapidly as one moves away from that centre in either direction. The resulting density curve is the familiar symmetric bell shape: it rises to a single peak at the centre, falls away monotonically on both sides, and approaches the horizontal axis asymptotically without ever touching it. Because the total area beneath the curve equals one, probabilities are represented by areas under this curve over specified intervals.

The distribution is completely determined by two parameters: a location parameter $\mu$, which fixes the centre of symmetry and coincides with the mean, and a scale parameter $\sigma^2$, which fixes the spread. Changing $\mu$ shifts the entire curve horizontally without altering its shape; changing $\sigma$ stretches or compresses the curve vertically while preserving total area one, so a larger $\sigma$ produces a flatter, wider bell and a smaller $\sigma$ produces a taller, narrower bell.

The practical difficulty is that the density involves the constant $\pi$ and the exponential function, so areas cannot be obtained by elementary antiderivatives. The resolution is standardisation: every normal random variable is a linear rescaling of one canonical member of the family, the standard normal variable $Z$ with mean $0$ and variance $1$. All probability computations therefore reduce to reading areas under a single tabulated curve, after shifting the origin (subtracting $\mu$) and changing the scale (dividing by $\sigma$). This shift-and-scale device is the computational heart of the topic, and it is exactly the same logic that underlies coding techniques for computing moments of grouped data.

The moments of the normal distribution are equally important. The symmetry of the density forces every odd central moment to vanish, so the distribution has zero skewness, while the even central moments follow a clean recursion expressible through the variance. These facts underpin the moment generating function, which in turn yields the mean and variance with minimal computation and establishes that linear transformations of normal variables remain normal.

2. Mathematical Formalization & Core Definitions

> Definition (Normal random variable). A continuous random variable $X$ is said to have a normal distribution with parameters $\mu$ and $\sigma^2$, written $X\sim N(\mu,\sigma^2)$, if its probability density function is

$$
f(x)=\frac{1}{\sigma\sqrt{2\pi}}\exp\left[-\frac{1}{2}\left(\frac{x-\mu}{\sigma}\right)^{2}\right],\qquad -\infty<x<\infty,
$$

> where $\mu\in\mathbb{R}$ and $\sigma>0$.

The density satisfies the two defining conditions of a probability density function: $f(x)\ge 0$ for all real $x$, and $\int_{-\infty}^{\infty}f(x)\,dx=1$. The parameter $\mu$ is the mean and $\sigma^2$ is the variance; $\sigma$ is the standard deviation. The curve is symmetric about the vertical line $x=\mu$, so $f(\mu-t)=f(\mu+t)$ for every $t$, and the peak value $f(\mu)=1/(\sigma\sqrt{2\pi})$ is attained only at $x=\mu$.

> Definition (Standard normal variable). If $Z\sim N(0,1)$, then $Z$ is called the standard normal variable, with density

$$
\phi(z)=\frac{1}{\sqrt{2\pi}}e^{-z^{2}/2},\qquad -\infty<z<\infty,
$$

> and distribution function

$$
\Phi(z)=P(Z\le z)=\int_{-\infty}^{z}\phi(t)\,dt.
$$

The values $\Phi(z)$ are tabulated for $z\ge 0$; values for negative arguments are recovered from the symmetry identity $\Phi(-z)=1-\Phi(z)$.

> Definition (Standardisation). If $X\sim N(\mu,\sigma^2)$, the standardised variable is

$$
Z=\frac{X-\mu}{\sigma},
$$

> and $Z\sim N(0,1)$.

Standardisation is the composition of two operations: the shift of origin $X-\mu$, which centres the distribution at zero, and the change of scale $1/\sigma$, which rescales the spread to unity. Geometrically, the transformation moves the peak of the density to the origin and adjusts the width so that the inflection points of the curve sit at $z=\pm 1$.

> Definition (Moment generating function). For a random variable $X$, the moment generating function is $M_X(t)=E(e^{tX})$, defined for all $t$ in an open interval containing zero.

For the normal distribution the moment generating function is finite for every real $t$ and equals $M_X(t)=\exp\left(\mu t+\tfrac{1}{2}\sigma^{2}t^{2}\right)$.