Recognise repeated Bernoulli trials and evaluate exact probabilities.
Notes and key methods
For \(X\sim\operatorname{Bin}(n,p)\), \(P(X=r)=\binom nrp^r(1-p)^{n-r}\), \(E(X)=np\), and \(\operatorname{Var}(X)=np(1-p)\).
The trials must have two outcomes, a fixed number of trials, constant success probability and independence. Translate words such as fewer than, at most and at least into integer limits before summing.
Check the conditions
n is a fixed non-negative integer and 0≤p≤1.
A probability such as P(X≥2) includes X=2.
Three worked examples
Example 1 · Binomial probabilities
Independent trials have constant success probability \(\frac{2}{5}\). Let X count the successes in 7 trials.
Find P(X=2).
Find P(X at least 2).
Find E(X).
Find Var(X).
Show the reasoning
(a)
\(\binom{7}2p^2(1-p)^{5}=\frac{20412}{78125}\)
Answer: \(\frac{20412}{78125}\).
(b)
\(1-P(X=0)-P(X=1)=\frac{65732}{78125}\)
Answer: \(\frac{65732}{78125}\).
(c)
\(np=\frac{14}{5}\)
Answer: \(\frac{14}{5}\).
(d)
\(np(1-p)=\frac{42}{25}\)
Answer: \(\frac{42}{25}\).
Example 2 · Binomial probabilities
Independent trials have constant success probability \(\frac{1}{2}\). Let X count the successes in 8 trials.
Find P(X=2).
Find P(X at least 2).
Find E(X).
Find Var(X).
Show the reasoning
(a)
\(\binom{8}2p^2(1-p)^{6}=\frac{7}{64}\)
Answer: \(\frac{7}{64}\).
(b)
\(1-P(X=0)-P(X=1)=\frac{247}{256}\)
Answer: \(\frac{247}{256}\).
(c)
\(np=4\)
Answer: \(4\).
(d)
\(np(1-p)=2\)
Answer: \(2\).
Example 3 · Binomial probabilities
Independent trials have constant success probability \(\frac{1}{5}\). Let X count the successes in 9 trials.
Find P(X=2).
Find P(X at least 2).
Find E(X).
Find Var(X).
Show the reasoning
(a)
\(\binom{9}2p^2(1-p)^{7}=\frac{589824}{1953125}\)
Answer: \(\frac{589824}{1953125}\).
(b)
\(1-P(X=0)-P(X=1)=\frac{1101157}{1953125}\)
Answer: \(\frac{1101157}{1953125}\).
(c)
\(np=\frac{9}{5}\)
Answer: \(\frac{9}{5}\).
(d)
\(np(1-p)=\frac{36}{25}\)
Answer: \(\frac{36}{25}\).
Common mistakes
Do not confuse a binomial count with a geometric waiting time.
Check strict and inclusive endpoints before using a cumulative probability.