Sum of Squares of Standard Gaussian Random Variables has Chi-Squared Distribution
Jump to navigation
Jump to search
Theorem
Let $X_1, X_2, \ldots, X_n$ be independent random variables.
Let $X_i \sim \Gaussian 0 1$ for $1 \le i \le n$ where $\Gaussian 0 1$ is the standard Gaussian Distribution.
Then:
- $\ds \sum_{i \mathop = 1}^n X^2_i \sim \chi^2_n$
where $\chi^2_n$ is the chi-squared distribution with $n$ degrees of freedom.
Proof
By Square of Standard Gaussian Random Variable has Chi-Squared Distribution, we have:
- $X^2_i \sim \chi^2_1$
for $1 \le i \le n$.
So, by Sum of Chi-Squared Random Variables, we have:
- $\ds \sum_{i \mathop = 1}^n X^2_i \sim \chi^2_{1 + 1 + 1 \ldots} = \chi^2_n$
$\blacksquare$