11/21/2013

Random variables and probability distribution

Random variable
Random variable is defined as a variable which takes
numerical values determined by the outcome of a random
Distribution Theory experiment.

Discrete and continuous Random Variables Probability distributions

Random variables which can take only some isolated
values is known as discrete random variables.
The set of possible values of a random variable
Eg:- No. of children in a family, number of misprints together with their respective probabilities is
per page. called probability distribution.
A random variable, which can take any value in Probability distribution describes how the total
some interval of real numbers is known as probabilities are distributed under the possible
continuous random variables. values of a random variable.
Eg:- height, weight, age, BP, life -hour of light bulb,
time to failure of an electric equipment

3 4

1
 11/21/2013

Eg:- X- number of heads obtained in three tosses of a

coin Discrete Probability Distribution

X P(X =x)
If a random variable is a discrete variable, its probability distribution
0 P(X =0) = P(TTT) is called a discrete probability distribution.

=1/8 An example will make this clear. Suppose you flip a coin two times.
This simple statistical experiment can have four possible outcomes:
1 P(X =1) = 3/8 HH, HT, TH, and TT. Now, let the random variable X represent the
number of Heads that result from this experiment. The random
2 P(X = 2) = 3/8
variable X can only take on the values 0, 1, or 2, so it is a discrete
3 P(X = 3) = 1/8 random variable.

Eg: - Probability distribution of X, the number of diseased male goat with 3 male
Discrete probability distribution offspring is given by the following p.m.f. p(x):

The probability distribution of discrete random variable

is called discrete probability distribution, they are x 0 1 2 3 Total
described by probability mass function (p.m.f) and
denoted by p(x), that satisfies the properties p(x) p(0)=1/8 p(1)=3/8 p(2)=3/8 p(3)=1/8 Σ p (x ) = 1
x

(i) p(x) ≥ 0 The above p.m.f is represented graphically by the following figure.

(ii) Σp(x) = 1
1

where p(x) = P(X = x)= probability that the r.v X takes p.m.f
p(x)
the value x.

0 0 1 2 3
7 8
X→

2
 11/21/2013

The graph of the c.d.f for the above

Cumulative Distribution Function(c.d.f) Eg. is shown below.
F(X) = P(X ≤ x) , 0 ≤ F(x) ≤ 1
The following table gives the c.d.f of the r.v X, the number of diseased male
patients in families with 3 male offspring

x p(x) F(x) 1

0 1/8 1/8 F(x)

1 3/8 4/8 ↑
2 3/8 7/8
0 0 1 2 3
3 1/8 1
x→
Total 1 -
(Consider F(1) = P(X≤1) = P(1 or fewer patients)
= p(0) + p(1) = 1/8 + 3/8 = 4/8 =1/2. similarly for other values of x).
9 10

Continuous Probability Distributions

If a random variable is a continuous variable, its probability distribution Since the continuous random variable is defined over a continuous
is called a continuous probability distribution. range of values (called the domain of the variable), the graph of the
A continuous probability distribution differs from a discrete probability density function will also be continuous over that range.


distribution in several ways.




The probability that a continuous random variable will assume a The area bounded by the curve of the density function and the x-
particular value is zero. axis is equal to 1, when computed over the domain of the variable.


As a result, a continuous probability distribution cannot be

expressed in tabular form.

Instead, an equation or formula is used to describe a continuous
The probability that a random variable assumes a value between a
probability distribution. and b is equal to the area under the density function bounded by a
and b.
Most often, the equation used to describe a continuous probability
distribution is called a probability density function. Sometimes, it is
referred to as a density function, a PDF, or a pdf. For a continuous
probability distribution, the density function has the following
properties:

3
 11/21/2013

Probability density curve

The probability density function (p.d.f) of a continuous
random variable X denoted by f(x) is a function that P(a≤X ≤b)
satisfies
(i) f(x) ≥ 0
(ii) ∫x f(x)dx = 1 where f(x)dx = P(x < X <x + dx)
i.e: The total area under the density curve is one.
a b

13 14

Cumulative distribution
x function (c.d.f)
Probability density function F(x) = P(x ≤ x) = ∫ f (t )dt
−∞

Properties
P(x<X<x+dx)=f(x)dx (1) 0 ≤ F(x) ≤ 1

F (x1) ≤ F (x2) , if x1 ≤ x2 ie. Increasing

function of x.
f(x)

(2) F (-∞) = 0
x x+dx

(3) F (+∞) = 1

15 * P (a ≤ x ≤ b) = F(b) –16F (a)

4
 11/21/2013

Discrete probability distributions

Cumulative distribution function Binomial Probability Distribution:
To understand binomial distributions and binomial probability, it helps
to understand binomial experiments and some associated notation;
so we cover those topics first.

Binomial Experiment:
A binomial experiment (also known as a Bernoulli trial) is a statistical
experiment that has the following properties:
F(x)
•The experiment consists of n repeated trials.
•Each trial can result in just two possible outcomes. We call one of
-∞ x ∞ these outcomes a success and the other, a failure.
•The probability of success, denoted by P, is the same on every trial.
•The trials are independent; that is, the outcome on one trial does
not affect the outcome on other trials.
17 18

Notation
Consider the following statistical experiment. You flip a coin 2 times
and count the number of times the coin lands on heads. This is a The following notation is helpful, when we talk about binomial
binomial experiment because: probability.




The experiment consists of repeated trials. We flip a coin 2 times. x: The number of successes that result from the binomial


Each trial can result in just two possible outcomes - heads or tails. experiment.




The probability of success is constant - 0.5 on every trial. n: The number of trials in the binomial experiment.



P: The probability of success on an individual trial.
The trials are independent; that is, getting heads on one trial does

Q: The probability of failure on an individual trial. (This is equal to
not affect whether we get heads on other trials.
1 - P.)


b(x; n, P): Binomial probability - the probability that an n-trial
binomial experiment results in exactly x successes, when the
probability of success on an individual trial is P.


nCr: The number of combinations of n things, taken r at a time.

5
 11/21/2013

Binomial Distribution
Binomial Probability
A binomial random variable is the number of successes x in n repeated
trials of a binomial experiment. The probability distribution of a The binomial probability refers to the probability that a binomial
binomial random variable is called a binomial distribution (also known experiment results in exactly x successes. For example, in the above
as a Bernoulli distribution). table, we see that the binomial probability of getting exactly one head
in two coin flips is 0.50.

The binomial distribution has the following properties: Given x, n, and P, we can compute the binomial probability based on

the following formula:
The mean of the distribution (μx) is equal to n * P .
Binomial Formula. Suppose a binomial experiment consists of n trials
The variance (σ2x) is n * P * ( 1 - P ).

and results in x successes. If the probability of success on an individual

The standard deviation (σx) is
σ = n * P * (1 − P) trial is P, then the binomial probability is:
b(x; n, P) = nCx * Px * (1 - P)n - x

Example
Cumulative Binomial Probability
A cumulative binomial probability refers to the probability that the The probability that a student is accepted to a prestigious college is 0.3.
binomial random variable falls within a specified range (e.g., is greater If 5 students from the same school apply, what is the probability that at
than or equal to a stated lower limit and less than or equal to a stated most 2 are accepted?
upper limit).
Solution: To solve this problem, we compute 3 individual probabilities,
For example, we might be interested in the cumulative binomial
using the binomial formula. The sum of all these probabilities is the
probability of obtaining 45 or fewer heads in 100 tosses of a coin (see
answer we seek. Thus,
Example 1 below). This would be the sum of all these individual
binomial probabilities. b(x < 2; 5, 0.3) = b(x = 0; 5, 0.3) + b(x = 1; 5, 0.3) + b(x = 2; 5, 0.3)

b(x < 45; 100, 0.5) =

b(x = 0; 100, 0.5) + b(x = 1; 100, 0.5) + ... + b(x = 44; 100, 0.5) + b(x = 45; 5 5 5
100, 0.5) =   *(0.3)0 *(0.7)5 +   *(0.3)1 *(0.7) 4 +   *(0.3)2 *(0.7)3
 
0  
1  2
b(x < 2; 5, 0.3) = 0.1681 + 0.3601 + 0.3087
b(x < 2; 5, 0.3) = 0.8369

6
 11/21/2013

Note:-
A trial of a random experiment that can only Poisson Distribution
two possible outcomes (success and failure) is called a (S.D. Poisson(1781 – 1840) , published in 1837)
Bernoulli trial (Swiss mathematician Jacob Bernoulli Let x be a random variable following Poisson distribution with p.m.f
(1654 –1705)).
e− λ λ x
If it come with a Bernoulli trial to assign a value 1 p(x) = , x = 0,1,2,...
x! λ > 0
to one outcome and 0 to other, such r.v is known as
dichotomous or Bernoulli r.v with p.m.f 1 1 1
e = 1 + + + + ... = 2.718282
1! 2! 3!
p.m.f satisfies the following properties
p(x) = px q1-x, x = 0,1 0<p<1,
p+ q =1 1. p(x)≥ 0
e−λλx
A binomial experiment consists of a sequence of n 2. Σp(x) = 1,ie, Σ =1
independent and identical Bernoulli trials. x!
25 26

Attributes of a Poisson Experiment

Notation
A Poisson experiment is a statistical experiment that has the following
The following notation is helpful, when we talk about the Poisson
properties:
distribution.


The experiment results in outcomes that can be classified as

e: A constant equal to approximately 2.71828. (Actually, e is the
successes or failures.


The average number of successes that occurs in a specified region is base of the natural logarithm system.)

known. λ : The mean number of successes that occur in a specified region.

The probability that a success will occur is proportional to the size of x: The actual number of successes that occur in a specified region.


the region. cP(x; λ ): The Poisson probability that exactly x successes occur in
a Poisson experiment, when the mean number of successes is λ .


The probability that a success will occur in an extremely small region
is virtually zero.
The Poisson distribution has the following properties:
Note that the specified region could take many forms. For instance, it

could be a length, an area, a volume, a period of time, etc. The mean & variance of the distribution is equal to λ .

7
 11/21/2013

Q2. A fire station switch board receives an average of 0.9 call per
Eg:- minute. Find the probability that there is ,

1. No. of telephone calls received by a fire station switch (i) No call in a minute.
board during a given period of time. (ii) Two or more calls in a minute.

2. No. of defective leather footwear among the exported (iii) What is the average no. of call per minute?
footwears (iv) What is the S.D of the no. of calls received per minutes?

3. No. of E-coli bacteria per small volume of urine. Ans. Let X be the no. of calls per minutes. Then X ∼ P(λ=0.9)

4. No. of leaks in a hundred km oil pipe line. e − .9 (0.9) 0

1.
P(X = 0) = = e − .9
5. No. of babies born blind per year in a city 0!
= 0.4066.
2. P (X≥2) = 1-P (X<2) = 1-[P (X=0) + P (X =1)]
Result: Binomial distribution tends to Poisson
e −.9 (0.9) 1
distribution when = 1 − 0 .4066 −
1!
n → ∞ and p → 0 such that np = λ is finite. = 0.2275
29 3. The average no. of calls received per minute, = 0.9 30

4. S.D = σ = λ = 0.9 = 0.95

Normal Distribution
Normal distribution represents the distributions of
continuous variables like height, weight, BP, experimental An important use of normal distribution is in
errors in scientific measurements etc. and this distribution approximating discrete distributions, like binomial
is often used to approximate other distributions. distribution for large values of ‘n’. It plays an important
role in statistical inference and in statistical quality
It provides a reasonable approximation to the distribution control.
of many different variables in the field of science and
engineering, business and commerce, biology, education,
sociology, agriculture etc.

31 32

8
 11/21/2013

A continuous random variable X is said to have a Normal distribution curve is symmetric and Bell shaped. Any
Normal distribution with parameters µ and σ, if its change in the parameter µ merely shift the curve to the left or
p.d.f is given by right, but any change in the parameter σ will changes the shape
1 1
− 2 ( x − µ )2
of the curve. ie, As σ increases the curve becomes more and
f (x ) = e 2σ more flat. The mean µ describes where the corresponding curve
2π σ is centered, and S.D σ describes how much the curve spreads out
= 0, O.W around the center.
-∞ < X < +∞
-∞ < µ < +∞, σ > 0
Probability Distribution
No Notationaly, X ∼ N(µ, σ2). Since f (x) is a p.d.f, it is to be noted that

(i) f(x)≥0 ∀ x and

∞
(ii)
∫ f (x ) dx = 1 -∞ x=µ +∞
−∞

33 34

x
Properties.
Cumulative density function (c.d.f) F(x) = ∫ f (x) dx (i) Normal distribution is bell shaped and is symmetric about its mean µ
−∞
=P(X≤x)
(ii) For a normal distribution mean = median = mode.
Cumulative Probability Distribution
(iii) The points of inflection of a normal density curve are at x= µ ± σ.
(iv) The ordinate of a normal density is maximum at x=µ and maximum
ordinate is
F(x)
1
2π σ
-∞ x +∞ (v) Almost all values of a normal variable lies in (µ-3σ,µ+3σ).
1
Note:- (i) F (µ ) =
2 For a normal distribution,
(ii) F( ∞ ) = 1

(iii) F(−∞) = 0 P( µ -σ <X<µ +σ ) = 0.6827

P( µ -2σ <X<µ +2σ) = 0.9545
P( µ -3σ <X<µ +3σ) = 0.9973
35 36
P( µ -6σ <X<µ +6σ) = 0.9999

9
 11/21/2013

Standard Normal Distribution c.d.f of z is given by

z z −t2
1
The normal distribution having Mean 0 and variance 1 is called standard normal ϕ (z) = P(Z ≤ z) = ∫ φ (t )dt = ∫ e 2
dt
distribution. It is usually denoted by the variable Z. ie. If X ∼ N(µ,σ2),then −∞ −∞ 2π
X −µ
Z = ∼ N(0,1).
σ
ϕ(t)
The probability density function of a standard normal variate Z is given by

−z2
1 -∞ t 0 z +∞
φ (z ) = e 2
, −∞ < Z < +∞
2π
Standard normal distribution possesses the same properties as the general
normal distribution does .

-∞ 37 z=0 38
+∞

Eg:- The systolic blood pressure X (in mm Hg) of an individual

Properties selected at random from a certain population is normally
distributed with mean of 120 mm Hg and S.D of 10 mm Hg.
(i) ϕ (0)= ½ Find the probability that the blood pressure of a randomly
(ii) ϕ (-∞)= 0 selected person from the said population will be
(iii) ϕ (∞)= 1
(iv) ϕ (-t) = 1- ϕ (t), since ϕ (-t) = P(Z ≤ -t) = P(Z≥ t) , (iii) Between 110 and 140 (iv) Below 90 or above 140
(due to symmetry of standard normal curve) • Below 90 (ii) Above 140
Sketch the areas under the respective density curves or each
= 1- P (Z ≤ t) = 1-ϕ (t)
cases.
If 10,000 persons selected at random from that population how
many would you expect to have S.B.P between 110mm Hg and
The above relationship is useful in determining different normal
probabilities.
140 mm Hg. Given ϕ(1) = 0.8413, ϕ(2) = 0.9772, ϕ(3) = 0.9987

X −µ
Ans. X ∼ N (µ=120, σ = 10), we know that Z = σ ∼ N(0,1).
39 40

10
 11/21/2013

110−120 140−120
 x - µ 90 − µ   90 − 120  (iii) P(110 < X <140) = P < Z< 
i) P(X<90) = P <  = P Z <   10 10 
 σ σ   10 
= P(−1 < Z < 2) = ϕ (2) − ϕ (−1)
= P(Z < −3) = ϕ (− 3) = 1 − ϕ (3)
= ϕ (2) − (1 − ϕ (1))
= 1 − 0.9987 = 0.0013 = 0.9772 –1 +0.8413 = 0.8185
90 120
x

110 120 140

X
(iv) P(X<90 or X>140) = P(X<90) + P(X>140)
= 0.0013+ 0.0228 (from (i) and (ii))
= 0.0241
-3 Z=0 3 z

 140 − 120 
(ii) P(X>140) = P Z >  = P(Z > 2)
 10 
= 1 − P (Z < 2 ) = 1 − ϕ (2)

= 1 − 0 . 9772 = 0 . 0228 90 120 140

120 140 x X
41 42

Example: An average light bulb manufactured by the Acme

Let Y be the no. of individuals having S.B.P between 110 Corporation lasts 300 days with a standard deviation of 50
and 140mm Hg, among 10,000 persons selected at days. Assuming that bulb life is normally distributed, what is
random from the said population. Then Y ∼ B (n=10,000,
the probability that an Acme light bulb will last at most 365
p = 0.8185).
Thus expected no. of individuals having S.B.P between days?
110 and 140mm Hg is E(Y) = n.p = Solution: Given a mean score of 300 days and a standard
10,000 * 0.8185 =8185. deviation of 50 days, we want to find the cumulative
probability that bulb life is less than or equal to 365 days.
Thus, we know the following:


The value of the normal random variable is 365 days.


The mean is equal to 300 days.


The standard deviation is equal to 50 days.
The answer is: P( X < 365) = 0.90. Hence, there is a 90%
43
chance that a light bulb will burn out within 365 days.

11
 11/21/2013

But sample sizes are sometimes small, and often we do not know the
Sampling Distributions standard deviation of the population. When either of these
problems occur, statisticians rely on the distribution of the t statistic
Student's t Distribution
(also known as the t score), whose values are given by:
The t distribution (Student’s t-distribution) is a probability
distribution that is used to estimate population parameters when the t = [ x - μ ] / [ s / sqrt( n ) ]
sample size is small and/or when the population variance is where x is the sample mean, μ is the population mean, s is the
unknown.
standard deviation of the sample, and n is the sample size. The
distribution of the t statistic is called the t distribution or the
Why Use the t Distribution? Student t distribution.
According to the central limit theorem, the sampling distribution of a The t distribution allows us to conduct statistical analyses on certain
statistic (like a sample mean) will follow a normal distribution, as long data sets that are not appropriate for analysis, using the normal
as the sample size is sufficiently large. Therefore, when we know the
distribution.
standard deviation of the population, we can compute a z-score, and
use the normal distribution to evaluate probabilities with the sample
mean.

Degrees of Freedom When to Use the t Distribution

There are actually many different t distributions. The particular form The t distribution can be used with any statistic having a bell-
of the t distribution is determined by its degrees of freedom. The shaped distribution (i.e., approximately normal). The central limit
degrees of freedom refers to the number of independent theorem states that the sampling distribution of a statistic will be
observations in a set of data. normal or nearly normal, if any of the following conditions apply.

Properties of the t Distribution The population distribution is normal.

The sampling distribution is symmetric, unimodal, without
The t distribution has the following properties: outliers, and the sample size is 15 or less.




The mean of the distribution is equal to 0 . The sampling distribution is moderately skewed, unimodal,


The variance is equal to v / ( v - 2 ), where v is the degrees of without outliers, and the sample size is between 16 and 40.

freedom (see last section) and v > 2. The sample size is greater than 40, without outliers.
The variance is always greater than 1, although it is close to 1 when The t distribution should not be used with small samples from
there are many degrees of freedom. With infinite degrees of populations that are not approximately normal.
freedom, the t distribution is the same as the standard normal
distribution

12
 11/21/2013

Chi-Square Distribution
Probability and the Student t Distribution
The distribution of the chi-square statistic is called the chi-square
When a sample of size n is drawn from a distribution. In this lesson, we learn to compute the chi-square
population having a normal (or nearly normal) statistic and find the probability associated with the statistic.
distribution, the sample mean can be The Chi-Square Statistic
transformed into a t score, using the equation
Suppose we conduct the following statistical experiment.
presented at the beginning of this lesson. We
We select a random sample of size n from a normal population,
repeat that equation below:
having a standard deviation equal to σ. We find that the standard
t = [ x - μ ] / [ s / sqrt( n ) ] deviation in our sample is equal to s.
Given these data, we can define a statistic, called chi-square, using
where x is the sample mean, μ is the population
the following equation:
mean, s is the standard deviation of the sample,
n is the sample size, and degrees of freedom are Χ2 = [ ( n - 1 ) * s2 ] / σ2

equal to n - 1.

If we repeated this experiment an infinite number

of times, we could obtain a sampling distribution
for the chi-square statistic. The chi-square
distribution is defined by the following
probability density function:
The chi-square distribution has the following properties:
2− ( n / 2) 2 n2 −1 χ2
2

( χ ) e ; 0 < χ2 < ∞

pdf =
The mean of the distribution is equal to the number of degrees of
freedom: µ = n.
n/2

The variance is equal to two times the number of degrees of
freedom: σ2 = 2 * n


When the degrees of freedom are greater than or equal to 2, the
e is a constant equal to the base of the natural maximum value for Y occurs when Χ2 = n- 2.


As the degrees of freedom increase, the chi-square curve approaches
logarithm system (approximately 2.71828).
a normal distribution.

13
 11/21/2013

F Distribution
The steps required to compute an f statistic:
The F distribution is the probability distribution


associated with the f statistic. In this lesson, we show Select a random sample of size n1 from a
how to compute an f statistic and how to find normal population, having a standard
probabilities associated with specific f statistic values.
deviation equal to σ1.


The f Statistic Select an independent random sample of size
The f statistic, also known as an f value, is a random n2 from a normal population, having a
variable that has an F distribution. standard deviation equal to σ2.
The f statistic is the ratio of s12/σ12 and s22/σ22.

The following equivalent equations are commonly used to compute an f The F Distribution
statistic:
The distribution of all possible values of the f statistic is called an F
f = [ s12/σ12 ] / [ s22/σ22 ] distribution, with v1 = n1 - 1 and v2 = n2 - 1 degrees of freedom.
f = [ s12 * σ22 ] / [ s22 * σ12 ] The curve of the F distribution depends on the degrees of freedom, v1
and v2.
χ 2   χ22 
f =  1   v2 
 v1  When describing an F distribution, the number of degrees of freedom
associated with the standard deviation in the numerator of the f statistic
f =  χ12 * v2   χ 2 2 * v1  is always stated first.
Thus, f(5, 9) would refer to an F distribution with v1 = 5 and v2 = 9
where σ1 is the standard deviation of population 1, s1 is the standard degrees of freedom
deviation of the sample drawn from population 1, σ2 is the standard
The F distribution has the following properties:
deviation of population 2, s2 is the standard deviation of the sample drawn


from population 2, Χ21 is the chi-square statistic for the sample drawn The mean of the distribution is equal to v2 / ( v2 - 2 ) for v2 > 2.
The variance is equal to [ 2 * v22 * ( v1 + v1 - 2 ) ] / [ v1 * ( v2 - 2 )2 *

from population 1, v1 is the degrees of freedom for Χ21, Χ22 is the chi-
( v2 - 4 ) ] for v2 > 4.
square statistic for the sample drawn from population 2, and v2 is the
degrees of freedom for Χ22 . Note that degrees of freedom v1 = n1 - 1, and
degrees of freedom v2 = n2 - 1 .

14