Axioms of Probability
What is a Probability?
The goal of probability is to assign some number, $P(A)$, called the probability of event $A$, which will give a precise measure to the chance that $A$ will occur. In statistics, we draw a sample from a population, and give an estimate. So, you will be able to understand statistics more thoroughly and deeply if you first understand probabilities.
-
Start with an experiment that generates outcomes.
-
Organize all of the outcomes into a sample space, $S$.
-
Let A be some event contained in $S$. That is, $A$ is some collection of outcomes from the experiment.
What do we expect to be true of $P(A)$?
Axioms of Probability
-
Axiom 1
For any event $A, 0\le P(A)\le 1$.
-
Axiom 2
$P(S)=1$
-
Axiom 3
If $A_1,A_2,\ldots,A_n$ are a collection of $n$ mutually exclusive events (i.e. the intersection of any two is the empty set), then
-
Axiom 3 extended
More generally, if $A_1,A_2,\ldots$ is an infinite collection of mutually exclusive events, then
These three properties are called the Axioms of Probability and we can derive many results from them.
Example 1
Experiment: Flip a coin until the first tail appears. Let $0$ represent a head and $1$ a tail.
\[S=\{1,01,001,0001,\ldots\}\]Let $A_n$ represent the event of obtaining a tail on the $n^{th}$ flip, $A_n={00\ldots01}$. Find $P(A_1),P(A_2),P(A_5)$ and $P(A_n)$, where $n$ is a positive integer.
\[\begin{align} &P(A_1)=1/2\\ &P(A_2)=P(\{0,1\})=1/4\\ &P(A_5)=P(\{00001\})=1/2^5\\ &P(A_n)=1/2^n \end{align}\]Note:
\[P(S)=P(\cup_{k=1}^{\infty}A_k)=\sum_{k=1}^{\infty}P(A_k)=\sum_{k=1}^{\infty}\frac{1}{2^k}=1\]If $B$ is the event that it takes at least $3$ flips to obtain a tail, find $P(B)$.
\[P(B)=P(\{001,0001,\ldots\})\]$B^c$, the complement of $B$, is the event that you obtain a tail on the first or second flip.
\[P(B^c)=P(\{1,01\})=1/2+1/4=3/4\]We also note:
\[\begin{align} &P(S)=P(B\cup B^c)=P(B)+P(B^c)=1.\text{ So,}\\ &P(B)=1-P(B^c)=1-3/4=1/4\\ \end{align}\]Consequenses of the Axioms
If $A$ and $B$ are two events contained in the same sample space $S$,
-
$A\cap A^c=0$ and $A\cup A^c=S$ so,
$1 = P(S)=P(A\cup A^c)= P(A)+P(A^c)$ which implies
$P(A^c)=1-P(A)$.
-
If $A\cap B=0$ then $P(A\cap B)=0$.
-
$P(A\cup B) = P(A) + P(B) - P(A\cap B)$.
These three consequenses will help us calculate many probabilities.
Example 2
Select a car coming off an assembly line and inspect it for 3 different defects (engine problem, seat belt problem, bad paint job).
\[\begin{align} S&=\{000,100,010,001,110,101,011,111\}\\ \mid S\mid &=8\\ \end{align}\]Consider the three events:
-
$A$ is the event defect 1 is present,
$A={100,110,101,111}$
-
$B$ is the event defect 2 is present,
$B={010,110,011,111}$
-
$C$ is the event defect 3 is present,
$C={001,011,101,111}$
Suppose over many days, data is collected and it is found that 20% of the cars have defect 1, 25% have defect 2, and 30% have defect 3. Further, 5% have defects 1 and 2, 7.5% have defects 2 and 3, 6% have defects 1 and 3, and 1.5% have all three.
\[\begin{align} &P(A)=0.2&&P(A\cap B)=0.05\\ &P(B)=0.25&&P(B\cap C)=0.75\\ &P(C)=0.3&&P(A\cap C)=0.06\\ &&&P(A\cap B\cap C)=0.15\\ \end{align}\]Calculate the probability of each the following events for the randomly selected car:
-
Defect 1 did not occur.
-
At least one defect occurs.
-
No defect occurs.
-
Defect 1 and 3 occur but 2 does not.
Answer:
-
$P(A^c) = 1 - P(A) = 1 - 0.2 = 0.8$
-
$P(\text{At least 1 defect})=P(A\cup B\cup C)$
$=P(A)+P(B)+P(C)-P(A\cap B)-P(B\cap C)-P(A\cap C)+P(A\cap B\cap C)$
$=0.2+0.25+0.3-0.05-0.75-0.06+0.15=0.58$
$P(\text{no defect})=P(A\cup B\cup C)^c=1-P(A\cup B\cup C)$
$=1-0.58=0.42$
- $P({101})=P(A\cap C)-P(A\cap B\cap C)=0.06-0.15=0.45$