Section 8.

1 – Probability Densities
ENGR 3311 – Engineering Math Methods
Instructor: Michael Weeks, PhD
Fall 2023
Continuous Random Variables

• A continuous random variable has a probability of 0 of

assuming exactly any of its values.
• Consequently, its probability distribution cannot be given in
tabular form.
• Consider a random variable whose values are the heights of
all students.
• Between any two values, say 173.5cm and 174.5cm there are
an infinite number of heights, one of which is 174cm.
• The probability of selecting a student at random who is
exactly 174cm tall and not one of the infinitely large set of
heights so close to 174cm is 0. (Not humanly possible)
• However, we can instead consider the probability of selecting
a person who is at least 173cm, or taller than 175cm.
• Now we are dealing with an interval (or continuum of points)
rather than a point value of our random variable.
• To compute the probability that a given random variable 𝑿 will
take on a value on the interval 𝒂, 𝒃 , such that 𝒂 𝒃, we divide
the interval into 𝒎 equal subintervals of width ∆𝒙.

a Δx b

x1 x2 x3 xm
• The probability that the random variable will take on a value in
the subinterval containing 𝒙𝒊 is given by 𝒇 𝒙𝒊 . ∆𝒙.
• Hence, the probability that the random variable 𝑿 will take on a
value in the interval 𝒂, 𝒃 is given by
𝒎

𝑷 𝒂 𝑿 𝒃 𝒇 𝒙𝒊 . ∆𝒙
𝒊 𝟏

2
Continuous Random Variables
• Although the probability distribution of a continuous
random variable cannot be presented in tabular form, it
can be stated as a formula.
• If 𝒇 𝒙 is an integrable function defined for all values of
the random variable, we define the probability that the
value of the random variable falls in 𝒂, 𝒃 , by letting
∆𝒙 → 𝟎, as

𝒃
𝑷 𝒂 𝑿 𝒃 𝒂
𝒇 𝒙 . 𝒅𝒙

Probability Density Functions

• The function 𝒇 𝒙 is denoted as the probability density function
(pdf) for the continuous random variables 𝑿, defined over the set
of real numbers, if

• 𝒇 𝒙 𝟎, for all 𝒙 ∈ 𝑹

• 𝒇 𝒙 . 𝒅𝒙 𝟏

• Hence, 𝒇 𝒙 cannot just be any real-valued integrable function.

• The cumulative distribution function:
𝒙
𝑭 𝒙 𝑷 𝑿 𝒙 𝒇 𝒕 . 𝒅𝒕

𝑷 𝒂 𝑿 𝒃 𝑭 𝒃 𝑭 𝒂
3
Probability Density Functions (Example 1)
If a random variable has the probability density
𝟐𝒆 𝟐𝒙
𝒇 𝒙 for 𝒙 𝟎
𝟎 for 𝒙 𝟎

find the probabilities that it will take on a value

a) between 1 and 3;
b) greater than 0.5.

4
Moments of a Probability Density

• For the first moment about the origin, the mean of a

probability density is,

𝝁 𝒙. 𝒇 𝒙 . 𝒅𝒙

• For the 2nd moment about the mean, the variance of a

probability density is,

𝝈𝟐 𝒙 𝝁 𝟐 . 𝒇 𝒙 . 𝒅𝒙 𝒙𝟐 . 𝒇 𝒙 . 𝒅𝒙 𝝁𝟐

Probability Density Functions (Example 3)

Determine the mean and variance of the probability

density given by the following function.

𝒇 𝒙 𝟐𝒆 𝟐𝒙 for 𝒙 𝟎
𝟎 for 𝒙 𝟎

7
Moments of a Probability Density

• The kth moment about the origin of a probability density is,

𝝁𝒌 𝒙𝒌 . 𝒇 𝒙 . 𝒅𝒙

• Hence, the mean of a probability density is,

𝝁 𝒙. 𝒇 𝒙 . 𝒅𝒙

• The 2nd moment about the mean (or variance) of a

probability density is,

𝝈𝟐 𝒙 𝝁 𝟐 . 𝒇 𝒙 . 𝒅𝒙 𝒙𝟐 . 𝒇 𝒙 . 𝒅𝒙 𝝁𝟐

Probability Density Functions (Example 3)

Determine the mean and variance of the probability

density given by the following function.

𝒇 𝒙 𝟐𝒆 𝟐𝒙 for 𝒙 𝟎
𝟎 for 𝒙 𝟎

8
The Normal Distribution

• The most important continuous probability distribution in

the entire field of statistics is the normal distribution.
• Its pdf approximately describes many phenomena that occur
in nature, industry, and research.
• The density of the normal random variable 𝑿, with mean 𝝁
and variance 𝝈𝟐 , is

𝟏 𝟏
𝒙 𝝁 𝟐
𝒇 𝒙; 𝝁, 𝝈 𝒆 𝟐𝝈 𝟐
∞ 𝒙 ∞
𝝈 𝟐𝝅

• The mathematical equation for the probability distribution

of the normal variable depends on the two parameters 𝝁
and 𝝈, its mean and standard deviation, respectively.
𝟏 𝟏
𝒙 𝝁 𝟐
𝒇 𝒙; 𝝁, 𝝈 𝒆 𝟐𝝈 𝟐

𝝈 𝟐𝝅

• The normal distribution has a bell-shaped curve which is

completely determined by the two parameters 𝝁 and 𝝈.

9
The Normal Distribution Characteristics
• The mode, which is the point on the horizontal axis where
the curve is a maximum, occurs at 𝒙 𝝁.
• The curve is symmetric about a vertical axis through the
mean 𝝁.
• The curve with the larger standard deviation is lower and
spreads out farther.
• The total area under the curve and above the horizontal
axis is equal to 𝟏.

• Probabilities relating to normal distributions are obtained

from special tables that pertain to the standard normal
distribution, namely the normal distribution with 𝝁 𝟎
and 𝝈 𝟏.
𝒛
𝟏 𝒕𝟐
𝑭 𝒛 𝑷 𝒁 𝒛 𝒆 𝟐 . 𝒅𝒕
𝟐𝝅

10
The Normal Distribution
• The probability that a random variable 𝒁 having the
standard normal distribution will take on a value between
𝒂 and 𝒃, is

𝑷 𝒂 𝒁 𝒃 𝑭 𝒃 𝑭 𝒂

The Normal Distribution (Example 4)

Find the probabilities that a random variable having the
standard normal distribution will take on a value
a) between 0.87 and 1.28;
b) between -0.34 and 0.62;
c) greater than 0.85;
d) greater than -0.65.

11
13
The Normal Distribution
• To compute the probabilities from the normal distribution, we
apply the following transformation to convert a normal random
variable 𝑿 into a standard normal random variable 𝒁, or its z-
score.

𝑿 𝝁
𝒁
𝝈

• We can then use the table for the standard normal distribution to
calculate the probabilities for the observations of any normal
random variable 𝑿.
𝒃 𝝁 𝒂 𝝁
𝑷 𝒂 𝑿 𝒃 𝑭 𝑭
𝝈 𝝈

The Normal Distribution (Example 5)

Given a random variable 𝑿 having a normal distribution with 𝝁 𝟓𝟎 and
𝝈 𝟏𝟎, find the probability that 𝑿 assumes a value between 𝟒𝟓 and 𝟔𝟐.

The Normal Distribution (Example 6)

Given a random variable 𝑿 having a normal distribution with 𝝁 𝟑𝟎𝟎 and
𝝈 𝟓𝟎, find the probability that 𝑿 assumes a value greater than 𝟑𝟔𝟐.

14
The Normal Distribution

• There are also problems in which we are given probabilities

relating to standard normal distributions and asked to find
the corresponding values of 𝒛.
• Let 𝒛𝜶 be such that the probability is 𝜶 that it will be
exceeded by a random variable having the standard
normal distribution.
𝜶 𝑷 𝒁 𝒛𝜶

NOTE: 𝑭 𝒛𝜶 𝑷 𝒁 𝒛𝜶 𝟏 𝜶

The Normal Distribution (Example 7)

Given a standard normal distribution, find the value of 𝒌

such that:
a 𝑷 𝒁 𝒌 𝟎. 𝟑𝟎𝟏𝟓, and
b 𝑷 𝒌 𝒁 𝟎. 𝟏𝟖 𝟎. 𝟒𝟏𝟗𝟕

16
Approximation to Binomial Distribution

• The normal distribution is often a good approximation to a

discrete distribution when the latter takes on a symmetric
bell shape.
• The normal distribution can be used to approximate the
binomial distribution when 𝒏 is large and 𝒑, the probability
of a success, is close to 𝟎. 𝟓.

• If 𝑿 is a random variable having the binomial distribution

with the parameters 𝒏 and 𝒑, the limiting form of the
distribution function of the standardized random variable
𝑿 𝒏𝒑
𝒁
𝒏𝒑 𝟏 𝒑

as 𝒏 → ∞ and 𝒑 is close to 𝟎. 𝟓, is given by the standard

normal distribution 𝒛
𝟏 𝒕𝟐
𝑭 𝒛 𝒆 𝟐 . 𝒅𝒕
𝟐𝝅

• Use the normal approximation to the binomial distribution

only when 𝒏𝒑 and 𝒏 𝟏 𝒑 are both greater than 𝟏𝟓.

18
Approximation to Binomial (Example 8)

The probability that a patient recovers from a rare blood

disease is 𝟎. 𝟒. If 𝟏𝟎𝟎 people are known to have contracted
this disease, what is the probability that:
a) more than 30 survive?
b) exactly 30 survive? Note that for this we consider
𝑷 𝑿 𝟑𝟎 𝑷 𝟐𝟗. 𝟓 𝑿 𝟑𝟎. 𝟓 .

• Let the binomial variable 𝑿 represent the number of patients who

survive. Since 𝒏 𝟏𝟎𝟎, we should obtain fairly accurate results
using the normal-curve approximation with

19
Checking if the Data Are Normal

• Many commonly used statistical procedures require that the

probability distribution be nearly normal.
• Hence, need a method to check if the data from the
experiment are normal.
• A graphical procedure can be used for detecting how close
(or far) the data is from normality. This will involve
subjective judgement.
• One graphical method for detecting departures from
normality is the normal scores plot, or normal quantile plot.

• Consider the case of the following data with a sample size of

4.
𝟔𝟕 , 𝟒𝟖 , 𝟕𝟔 , 𝟖𝟏
• To construct a normal scores plot,
1) Order the data from smallest to largest 𝒅𝟏 , 𝒅𝟐 , ⋯ , 𝒅𝒏 ;
2) Obtain the normal scores 𝒎𝟏 , 𝒎𝟐 , ⋯ , 𝒎𝒏 ;
3) Plot the kth observation 𝒅𝒌 versus the kth normal score
𝒎𝒌 , for all k.
• The normal scores refers to an idealized sample from the
standard normal distribution.
• It consists of the values of 𝒛 that divide the axes into
equal probability intervals.
𝒎𝟏 𝒛𝟎.𝟐 𝟎. 𝟖𝟒
𝒎𝟐 𝒛𝟎.𝟒 𝟎. 𝟐𝟓
𝒎𝟑 𝒛𝟎.𝟔 𝟎. 𝟐𝟓
𝒎𝟒 𝒛𝟎.𝟖 𝟎. 𝟖𝟒

21
Checking if the Data Are Normal

𝒅𝒌 𝟒𝟖 , 𝟔𝟕 , 𝟕𝟔 , 𝟖𝟏
𝒎𝒌 𝟎. 𝟖𝟒 , 𝟎. 𝟐𝟓 , 𝟎. 𝟐𝟓 , 𝟎. 𝟖𝟒

• The following is the plot of the ordered observations versus

the normal scores, i.e. 𝒎𝒌 , 𝒅𝒌 .

• Recall the following is the relationship between the

standard normal random variable 𝒁 and normal random
variable 𝑿.
𝒙 𝝁
𝒛 → 𝒙 𝝁 𝝈𝒛
𝝈
• If the values for 𝒙 were normal, it will have a linear
relationship with 𝒛.

22
Checking if the Data Are Normal

• The closer the appearance of a normal scores plot to a

straight line, the more plausible the data is from a normal
distribution.

• In practice, a minimum of 15–20 observations is needed in

order to evaluate the agreement with normality.
• With 𝒏 observations, the set of normal scores can be
evaluated as follows.
𝒎𝟏 𝒛 𝜶𝟏
⋮ ⋮
𝒎𝒌 𝒛 𝜶𝒌
⋮ ⋮
𝒎𝒏 𝒛 𝜶𝒏

𝒌
where 𝜶𝒌 𝑷𝒓𝒐𝒃 𝒁 𝒛𝜶𝒌 and 𝜶𝒌 .
𝒏 𝟏

23