The t distribution is another that is closely associated with random samples from a normal distribution.
Definition
If \(Z \sim \NormalDistn(0, 1)\) and \(Y \sim \ChiSqrDistn(k \text{ df})\) are independent random variables, then
\[ T \;\;=\;\; \frac{Z}{\sqrt{\diagfrac{Y}{k}}} \]has a distribution called a t distribution with k degrees of freedom,
\[ T \;\;\sim\;\; \TDistn(k \text{ df}) \]We cannot derive the pdf of the t distribution here, but simply state it without proof.
Probability density function of t distribution
If a random variable has a t distribution with \(k\) degrees of freedom then its probability density function is
\[ f(x) \;\;=\;\; \frac{\Gamma(\frac{k+1}{2})}{\Gamma(\frac{k}{2}) \sqrt{k\pi}} \left(1 + \frac{x^2}{k} \right)^{-\frac {k+1}{2}} \]Probabilities and quantiles
Probabilities and quantiles for t distributions must be evaluated numerically. In Excel, you can find \(P(Y \lt \langle y \rangle)\) for a t distribution with \(\langle df \rangle\) degrees of freedom with
=T.DIST(\(\langle y \rangle\), \(\langle df \rangle\), TRUE)
In a similar way, its \(\langle p \rangle\)'th quantile is
=T.INV(\(\langle p \rangle\), \(\langle df \rangle\))