Expectation of Negative Binomial Distribution/First Form

From ProofWiki
Jump to navigation Jump to search

Theorem

Let $X$ be a discrete random variable with the negative binomial distribution (first form) with parameters $n$ and $p$.


Then the expectation of $X$ is given by:

$\expect X = \dfrac {n p} q$

where $q = 1 - p$.


Proof

From Probability Generating Function of Negative Binomial Distribution (First Form):

$\map {\Pi_X} s = \paren {\dfrac q {1 - p s} }^n$


From Expectation of Discrete Random Variable from PGF:

$\expect X = \map {\Pi'_X} 1$


We have:

\(\ds \map {\Pi'_X} s\) \(=\) \(\ds \map {\frac \d {\d s} } {\paren {\dfrac q {1 - p s} }^n}\)
\(\ds \) \(=\) \(\ds \dfrac {n p} q \paren {\dfrac q {1 - p s} }^{n + 1}\) First Derivative of PGF of Negative Binomial Distribution/First Form


Plugging in $s = 1$:

\(\ds \map {\Pi'_X} 1\) \(=\) \(\ds \frac {n p} q \paren {\frac q {1 - p} }^{n + 1}\)
\(\ds \) \(=\) \(\ds \frac {n p} q \paren {\frac q q}^{n + 1}\) as $p = 1 - q$
\(\ds \leadsto \ \ \) \(\ds \expect X\) \(=\) \(\ds \frac {n p} q\) simplifying

$\blacksquare$


Sources