Scribd
Upload
110 views

LectureNotes9 PDF

This document discusses continuous random variables and the normal distribution. It begins by introducing continuous random variables and the standard normal distribution. It then discusses the general normal distribution by defining its probability density function and describing how it relates to the standard normal. Several properties of the normal distribution are provided, including how to calculate probabilities and quantiles. Examples are given to demonstrate calculating probabilities and finding critical values for the normal distribution.

Uploaded by

sharief85
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
110 views

LectureNotes9 PDF

This document discusses continuous random variables and the normal distribution. It begins by introducing continuous random variables and the standard normal distribution. It then discusses the general normal distribution by defining its probability density function and describing how it relates to the standard normal. Several properties of the normal distribution are provided, including how to calculate probabilities and quantiles. Examples are given to demonstrate calculating probabilities and finding critical values for the normal distribution.

Uploaded by

sharief85
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 21

Continuous Random Variable and

Their Probability Distributions:


Part III
Cyr Emile MLAN, Ph.D.
[email protected]
Continuous Random Variables: Part III

p. 1/21

Introduction
Text Reference: Introduction to Probability and Its
Application, Chapter 5.
Reading Assignment: Sections 5.6, March 30 April 1

In this set of notes, we will discuss the most important


and most widely used distribution in statistical
application.
In fact, many numerical measurements (heights,
weights, reaction time in psychological experiments,
intelligence and aptitude scores) have distributions that
can be approximated by a normal curve.

In addition, whenever an average of random variables is


computed, it tends to be spread out like the variation in
Continuous Random Variables: Part III
p. 2/21
a normal curve.

The Standard Normal Distribution


Density and Distribution Function
Let Z be a continuous random variable with density
1 z2
f (z) = e 2 , < z < .
2

Then Z has a standard normal distribution, denoted by


Z N (0, 1).

Sometimes, we also use (z) to denote the probability


density function of Z , i.e.,
1 z2
(z) = e 2 , < z < .
2
Continuous Random Variables: Part III

p. 3/21

The Standard Normal Distribution


The corresponding cumulative density
function of Z is given by
(z) = P (Z z) =

(u)du =

1 u2
e 2 du.
2

The standard normal distribution is symmetric about


z = 0, i.e.,
(z) = (z)
and
(z) = 1 (z).
(z) is tabulated on page 448 of textbook.

Continuous Random Variables: Part III

p. 4/21

The Standard Normal Distribution


Mean and Variance
We have
E[Z] = 0
Var(Z) = 1

Example 5.13:

If Z N (0, 1), then we have the following 3- rule:


P (1 Z 1) 0.68 = 68%,
P (2 Z 2) 0.95 = 95%,
P (3 Z 3) 0.997 = 99.7%.

Continuous Random Variables: Part III

p. 5/21

The General Normal Distribution


Density
A random variable X is said to be normal, if its density
is given by
(x)2
1
e 22 , < x < ,
f (x) =
2

where < < is a location parameter and > 0 is


a spread parameter.

Notation
X N (, 2 ).

Relationship with Standard Normal


X
X N (, ) iff Z =
N (0, 1).

Continuous Random Variables: Part III

p. 6/21

The General Normal Distribution

0.1

0.2

0.3

mu=0, sigma=1
mu=3, sigma=1
mu=0, sigma=2

0.0

probability density function

0.4

The Normal Distribution with parameter mu and sigma

0
2
4
6Part III
Continuous
Random Variables:
x

p. 7/21

Histogram Plots
Example 5.14: Quetelets data on 5738 Scottish

Soldiers.
Chest Size Frequency Chest Size Frequency
33
3
41
934
34
18
42
658
35
81
43
370
36
185
44
92
37
420
45
50
38
749
46
21
39
1073
47
4
40
1079
48
1
The histogram seems to indicate that the data is
normally distributed. Continuous Random Variables: Part III

p. 8/21

0.10
0.05
0.00

Relative Frequency

0.15

0.20

Histogram of Quetelets data on Scottish soldiers Chest Size

33

35

37

39

41

43

45

47

Continuous Random Variables: Part III


Chest Size

p. 9/21

Calculating Normal Probabilities


Distribution Function
If X N (, 2 ), then
P (X x)

X
x



x
= P Z



x
=
.

= P

Mean and Variance


We have
E(Y ) =

Var(Y ) = 2 .
Continuous Random Variables: Part III

p. 10/21

The General Normal Distribution


Example 5.15:
A firm that manufactures and bottles apple juice has a machine that
automatically fills bottles with 16 ounces of juice. The bottle can hold
up to 17 ounces. Over a long period, the average amount dispensed
into the bottle has been 16 ounces. However, there is variability in
how much juice is put in each bottle; the distribution of these
amounts has a standard deviation of 1 ounce. If the ounces of fill per
bottle can be assumed normally distributed, find the probability that
the machine will overflow any one bottle.

Solution:

Let X denote the amount of liquid (in ounces) dispensed into one
bottle by the filling machine. Then X N ( = 16, 2 = 1). A bottle
will overflow if the machine attempts to put more than 17 ounces in it.
Hence,


17 16
P (X > 17) = P Z >
= P (Z > 1) = 0.1587
1
Continuous Random Variables: Part III
p. 11/21

Joke

How is a normal distribution


like a lion. They both have
a mean mew!!!

Continuous Random Variables: Part III

p. 12/21

The General Normal Distribution


Quantiles of the Normal Distribution
Let 0 < < 1. Let Q, (1 ) denote the (1 )th
quantile of N (, 2 ). Then we have
P (Y Q, (1 )) = 1 .
Note that
P (X Q, (1 )) = P

Q, (1 )
X

Thus, we have
Q, (1 )
= Q0,1 (1 ) = 1 (1 )

where 1 (q) is the inverse function of the N (0, 1) cdf.

Hence
Q, (1 ) = + 1 (1 ),

Continuous Random Variables: Part III

p. 13/21

The General Normal Distribution


Example 5.16: Example 5.15 revisited.
Suppose it is decided to increases the size of each
bottle so that bottle overflown will happened less
frequently. Find the bottle size, x so that an x-ounce
bottle will overflow only 2.5% of the time.

Solution:
We must solve for x such that P (X > x) = 0.025 or
equivalently, P (X x) = 0.975. Hence,
x = + 1 (0.975) = 16 + 1(1.96) = 17.96

Continuous Random Variables: Part III

p. 14/21

The General Normal Distribution


Theorem 5.10:
Let a 6= 0 and b be two arbitrary constants and a random
variable X such as X N (, 2 ). Then the random variable
Y = aX + b is also normally distributed with parameters
E(Y ) = a + b
Var(Y ) = a2 2

Theorem 5.11 :
Let X1 N (1 , 12 ) and X2 N (2 , 22 ) be two independent random variables. We have
X1 + X2 N (1 + 2 , 12 + 22 )
X1 X2 N (1 2 , 12 + 22 )
Continuous Random Variables: Part III

p. 15/21

The General Normal Distribution


Example 5.17:

Data from the National Oceanic and Atmospheric


Administration indicate that the yearly precipitation in
Los Angeles is a normal random variable with a mean
of 12.08 inches and a standard deviation of 3.1 inches.
a). Find the probability that the total precipitation during
the next 2 years will exceed 25 inches.
b). Find the probability that next years precipitation will
exceed that of the following year by more than 3
inches. Assume that the precipitation totals for the
next 2 years are independent.

Continuous Random Variables: Part III

p. 16/21

The General Normal Distribution


Solution:
Let X1 and X2 be the precipitation totals for the next 2 years.
Because X1 and X2 are independent, we have

a.

X1 + X2

N (24.16, 19.22)

X1 X2

N (0, 19.22)

We have

P X1 + X2 > 25 =
=

b.

25 24.16
X1 + X2 24.16

>
19.22
19.22
P (Z > .19) = .4247
P

We have


X1 X2 0
30

>
19.22
19.22
= P (Z > .68) = 1 .7517 = .2483


P X1 X2 > 3 = P

Continuous Random Variables: Part III

p. 17/21

Joke

Dentist: "Good grief! Youve got the


biggest cavity Ive ever seen the
biggest cavity Ive ever seen"
"You dont have to repeat it, Doc!"
Dentist: "I didnt that was the echo."

Continuous Random Variables: Part III

p. 18/21

Quantile-Quantile Plot
The quantile-quantile plot or Q-Q plot is a simple (but
general) graphical display tool aimed at corroborating
whether or not observations from a dataset can be
(approximately) regarded as if they have been
generated by a given theoretical population distribution.
For that, the procedure compares the observed
quantiles of the data to the theoretical quantiles of the
probability distribution.
Consider n data points x1 , x2 , , xn . Order these data
points from the smallest to the largest to get the sorted
observations x(1) x(2) x(n) .


100i
Find yi =
percentiles of the theoretical
n+1
distribution.
Random Variables: Part III
Plot the pairs (yi , x(i) )Continuous
.

p. 19/21

Quantile-Quantile Plot
If the pairs (yi , x(i) ) lies on a straight line or
approximately, then we are confident that the theoretical
distribution can be regarded as the underlying
distribution that generated the observations
x1 , x2 , , xn .
If the pairs (yi , x(i) ) depart substantially from a linear
pattern, we claim that the observations were not
generated by the theoretical distribution.

When the theoretical distribution is the Normal


distribution, we will use the percentiles of the standard
normal distribution on the x-axis. In this case, we
regard the intercept as the mean and the slope as the
standard deviation . In practice, the line considered is
the one that goes through the .25 and .75 sample
quantiles and theoretical percentiles of the normal
Continuous Random Variables: Part III
p. 20/21
distribution.

Quantile-Quantile Plot

8
6
4
2
0
2

Sample Quantiles of 200 observations from Normal(3,4)

Normal QQ Plot: intercept=3.073 and sigma=2.185

Continuous Random Variables: Part III


Theoretical Quantiles

p. 21/21

You might also like