Continuous Random

Variables

Dr. Ma Liezl C. Gallardo

Defining Continuous Random
Variables
Exam Completion Time Fossil Age
The time it takes to complete a 60- Age of a fossil. Possible values
minute test. Possible values include include all real numbers between
all real numbers on the interval minimum and maximum age.
[0,60].

Vehicle Efficiency
Miles per gallon for a Toyota Prius. Possible values include all real numbers
between minimum and maximum MPG.

The key distinction between continuous and discrete random variables lies in how
probability is measured. For continuous variables, probability is measured over
intervals rather than at exact points. For instance, finding the probability of
completing an exam in exactly 32 minutes makes little sense, as it might take
32.012342472... minutes.
Probability Density Functions
Define the Function
A probability density function (pdf) helps find probabilities for
continuous random variables.

Ensure Non-Negative Range

The function must always have a non-negative output.

Calculate Area
Probability is determined by finding the area under the function curve.

Validate Total Area

To be valid, the total area under the curve must equal 1.

If a drawing represents a valid probability density function for a random variable X,

then P(a < X < b) equals the shaded area between points a and b under the curve.
Comparing Discrete vs.
Continuous Distributions
Discrete Distributions Continuous Distributions

Countable Uncountable

Discrete Points Continuous Intervals

Points have probability Points have no probability

p(x) is probability distribution function f(x) is probability density function

p(x) ≥ 0 f(x) ≥ 0

Σp(x) = 1 Total Area under curve = 1

This comparison highlights the fundamental differences between discrete and

continuous probability distributions. While discrete distributions assign probabilities to
specific points, continuous distributions spread probability across intervals, with
individual points having zero probability.
Example: Driving Time to School

0.20 0.45 0.80

P(X < 15) P(15 ≤ X ≤ 25) P(X > 15)
Chance of driving less than 15 minutes Chance of driving between 15 and 25 Chance of driving more than 15
minutes minutes

The time to drive to school for a community college student is an example of a continuous random variable. The probability
density function and areas of regions created by the points 15 and 25 minutes are shown in the graph. How long it takes a
student to drive to college is a good example of a continuous random variable. For continuous variables, single points like
"exactly 15 minutes" have no chance of happening. This is why P(X < 15) equals P(X ≤ 15). The total chance of driving more
than 15 minutes is 0.45 + 0.35 = 0.80.
Finding Percentiles in
Continuous Distributions
Define the Percentile
The pth percentile (x^p) is the value such that P(X < x^p) = p/100

Identify the Area

Locate the area under the curve that corresponds to the desired
percentile

Find the Value

Determine the x-value that creates that area from the left side of
the distribution

In our driving time example, we can identify that the 20th percentile is 15
minutes, meaning 20% of students take less than 15 minutes to drive to
school. Similarly, the 65th percentile is 25 minutes, indicating 65% of
students take less than 25 minutes for their commute.
Expected Value and
Variance

Expected Value Variance (σ²) Standard

(μ) Deviation (σ)
Population variance:
The population mean, σ² = Var(x) = E[(x- Population standard
denoted as μ = E(x) μ)²] deviation: σ = √Var(x)

The mean and variance can be calculated for most continuous random
variables, though the actual calculations require calculus beyond the
scope of this course. We use the same symbols to define expected
value and variance as we did for discrete random variables.
Applications and Special Distributions

Normal Distribution Uniform Distribution

The bell-shaped curve used for natural Equal probability across all values in an
phenomena, heights, weights, and test interval, used for random number
scores generation

Other Distributions
Exponential Distribution
Special distributions like chi-square, t-
Models time between events, such as
distribution, and F-distribution have
customer arrivals or equipment failures
important applications in statistics

These special continuous random variables have practical applications across various fields including engineering, finance,
quality control, and scientific research. Each distribution has unique properties that make it suitable for modeling different
real-world phenomena.
Exponential Distribution

The exponential distribution is a powerful statistical model used to analyze

waiting times until an event occurs. Whether you're waiting for a text
message or anticipating when equipment might fail, this distribution
provides valuable insights into random processes that occur at a constant
rate.
Key Properties of the
Exponential Distribution
Models Waiting Time
Used to model the time until an event occurs, such as receiving a text
message or experiencing equipment failure

Single Parameter Model

Defined by one parameter μ, which represents the expected waiting time

Memoryless Property
Future waiting time is independent of past waiting time - expressed as P(X >
a) = P(X > a+b | X > b)

Models "Breaking" Not "Wearing Out"

Appropriate for sudden failures rather than gradual deterioration over time
The Memoryless Property
What It Means Real-World Applications

The memoryless property is what makes the exponential This property applies to events that don't "wear out" or "age" -
distribution unique. If you've been waiting for 3 minutes for a the probability of occurrence remains constant regardless of
text message that typically arrives every 5 minutes, you still how much time has passed.
expect to wait another 5 minutes.
Examples include: radioactive decay, arrival of customers at a
Mathematically expressed as: P(X > a) = P(X > a+b | X > b) service point, and electronic component failures.
Example: Smartphone
Screen Cracking
Problem Statement Probability of Lasting
The time until a smartphone
600+ Hours
screen cracks follows an P(X > 600) = e^(-600/500) = e^(-
Exponential distribution with μ = 1.2) = 0.3012
500 hours of use.
There's about a 30% chance the
screen won't crack for at least
600 hours.

Median Cracking Time

x₅₀ = -μ ln(1-0.5) = -500 ln(0.5) = 347 hours

Half of all smartphone screens will crack before 347 hours of use.
The Memoryless Property in Action
New Phone
Expected lifetime: 500 hours

After 500 Hours of Use

Still expected to last another 500 hours

Probability Calculation
P(X > 1100 | X > 500) = P(X > 600) = 0.3012

This example demonstrates the memoryless property perfectly. If a phone has already lasted 500 hours without its screen cracking,
the probability that it will last an additional 600 hours is exactly the same as the probability that a brand new phone will last 600
hours. The phone doesn't "remember" that it's already been used for 500 hours.
Relationship with Poisson
Distribution
Poisson Process
Models number of occurrences in fixed time period

Mathematical Relationship
If events occur at rate μ, waiting time follows Exponential with mean
1/μ

Exponential Distribution
Models waiting time between occurrences

When events occur at a constant rate (a Poisson process), the waiting time between
consecutive events follows an exponential distribution. For example, if accidents
occur at a rate of 3 per month (Poisson with μ=3), then the waiting time until the next
accident follows an Exponential distribution with μ=1/3 month.
Example: Oil Refinery Accidents

3 1/3 0.9975
Accidents per Month Mean Waiting Time Probability
Constant rate following Poisson process In months between accidents Chance of waiting less than 2 months

For an oil refinery where accidents occur at a constant rate of 3 per month, we can calculate the probability of waiting less than 2 months
for the next accident: P(X < 2) = 1 - e^(-2/(1/3)) = 1 - e^(-6) = 0.9975. This means there's a 99.75% chance that another accident will occur
within the next 2 months.
Uniform Distribution: A Comparison

Flat Shape Two Boundaries

Constant chance across all values in the Set by lowest (a) and highest (b)
range possible values

Example: Tea Supply Same Likelihood

Tea amount between 50-1050g has Every value in the range has the same
equal chances at all levels chance of occurring

Unlike the exponential distribution, the uniform distribution gives the same chance to all values in a range.

For example, if someone buys 1000g of tea when they have 50g left, their tea supply will always be between 50g and 1050g. Any amount
in this range has the same chance of being the current supply.
Uniform Distribution

Uniform distribution is a fundamental concept in probability theory where all

values between a minimum and maximum have equal probability. This
presentation explores the key properties, formulas, and real-world
applications of uniform distributions.
Fundamentals of Uniform
Distribution
Definition Parameters
A continuous random variable where Defined by two values: minimum (a)
all values between a minimum (a) and maximum (b).
and maximum (b) have the same
probability.

Probability Density Function

f(x) = 1/(b-a), creating a rectangular shape on graphs.

The uniform distribution is one of the simplest continuous probability distributions. Its
rectangular shape makes it visually distinctive and mathematically straightforward to
work with. This distribution appears in many real-world scenarios where any value in
a range is equally likely to occur.
Key Statistical Measures
Expected Value (Mean)
μ = (a+b)/2

Variance
σ² = (b-a)²/12

Standard Deviation
σ = √[(b-a)²/12]

Percentile
xₚ = a + p(b-a)

These formulas allow us to calculate important statistical properties of any uniform distribution.
The expected value is simply the midpoint of the range, while the variance and standard
deviation measure the spread of possible values. The percentile formula enables us to find
specific points within the distribution.
Example: Loose Leaf Tea Inventory
A tea lover purchases Tie Guan Yin loose leaf tea in 1000g packages
when inventory reaches 50g.

The amount in stock follows a uniform distribution with a = 50g and

b = 1050g.

The probability density function is f(x) = 1/(1050-50) = 0.001, creating

a rectangular distribution.

50g 1050g
Minimum Stock Maximum Stock
Reorder point After purchase

550g 289g
Expected Value Standard Deviation
Average amount Spread measure
Calculating Probabilities
/

For the tea example, finding the probability of having at least 700g in stock is straightforward:

P(X ≥ 700) = (1050-700)/(1050-50) = 350/1000 = 0.35 or 35%.

This geometric approach makes probability calculations intuitive for uniform distributions.

Similarly, we can find percentiles using linear interpolation.

The 80th percentile would be x₈₀ = 50 + 0.80(1050-50) = 850g,
meaning 80% of the time, the tea stock is 850g or less.

Area = Probability Width × Height Simple Calculation

Use rectangle area (b-a) × f(x) P(X ≥ c) = (b-c)/(b-a)
Example: Waiting for a Train
Minimum Wait
0 minutes (just caught the train)

Expected Wait
10 minutes (mean waiting time)

Maximum Wait
20 minutes (just missed the train)

The Sounder commuter train arrives at Tacoma station every 20 minutes during rush
hour. The waiting time follows a uniform distribution with a = 0 and b = 20 minutes. The
expected waiting time is μ = (0+20)/2 = 10 minutes, with a standard deviation of σ = 5.77
minutes.

The probability of waiting at least 15 minutes is P(X ≥ 15) = (20-15)/(20-0) = 0.25 or 25%.
Quartiles and Interquartile Range
First Quartile (Q1)
x₂₅ = a + 0.25(b-a)

Second Quartile (Q2)

x₅₀ = a + 0.50(b-a) = median

Third Quartile (Q3)

x₇₅ = a + 0.75(b-a)

Interquartile Range (IQR)

IQR = Q3 - Q1

For the train example, Q1 = 0 + 0.25(20) = 5 minutes and Q3 = 0 + 0.75(20) = 15 minutes.

The interquartile range is IQR = Q3 - Q1 = 15 - 5 = 10 minutes, meaning the middle 50% of
waiting times fall within this 10-minute range.

Quartiles help us understand how values are distributed and provide a measure of spread
that is resistant to outliers.
Conditional Probabilities

Create New Distribution Restricted Range

Adjust parameters based on condition New minimum = condition value

Apply to Decision Making

Recalculate Probability
Inform choices with updated probabilities
Use standard formula with new range

For conditional probabilities with uniform distributions, we create a new distribution with adjusted parameters. For example, if a
commuter has already waited 5 minutes for a train, the conditional probability of waiting at least 10 more minutes becomes P(X ≥ 15
| X ≥ 5).

We can solve this by creating a new uniform distribution Y with minimum a = 5 and maximum b = 20, then calculating P(Y ≥ 15) =
(20-15)/(20-5) = 5/15 = 0.33 or 33%.
Normal Distribution
The most important probability distribution in Statistics is the Normal
Distribution, the iconic bell‐shaped curve. The Normal Distribution is
symmetric and defined by two parameters: the expected value (mean)
which describes the center of the distribution and the standard
deviation, which describes the spread.

The extremely complicated probability distribution function for the

Normal Distribution is:
Examples of Normal Distribution

There are many examples of data that are both symmetric and clustered towards the mean. For example, the dot
plots of the weights of apples and oranges. You can see that both graphs are clustered towards the center and
are symmetric. A Normal Distribution would be an appropriate model for the weight of apples and oranges.
Examples of Normal Distribution

There are many examples of data that are both symmetric and clustered towards the mean.

The Normal Distribution is defined by two parameters: the expected value (mean) μ, which describes the center of the
distribution, and the standard deviation σ, which represents the spread. Different values of these parameters create different-
shaped normal curves, all maintaining the characteristic bell shape.
Standard Normal Distribution
Definition Importance
A special case of the Normal The Standard Normal Distribution
Distribution is when μ = 0 and σ = serves as a reference distribution
1. This random variable is known that allows us to standardize any
as the Standard Normal normal random variable, making
Distribution and is always probability calculations more
represented by the letter Z. straightforward.

Z-Score Formula
Any Normal Random Variable X with Expected value μ and Standard
Deviation σ can be converted to a Standard Normal Distribution by using
the formula: Z = (X-μ)/σ
Probabilities in the Standard Normal Distribution

68% 95% 99.7%

Within 1σ Within 2σ Within 3σ
P(-1 < Z < 1) = 0.3413 + 0.3413 = 0.6826 P(-2 < Z < 2) = 0.3413 + 0.3413 + 0.1359 P(-3 < Z < 3) = 0.3413 + 0.3413 + 0.1359
+ 0.1359 = 0.9544 + 0.1359 + 0.0214 + 0.0214 = 0.9972

This means that for the standard Normal Distribution that 68% of the probability is between 1 and ‐1, 95% of the probability is
between ‐2 and 2 and 99.7% of the probability is between ‐3 and 3. These percentages may seem familiar from the Empirical Rule in
Chapter 3.
The Empirical Rule

The Empirical Rule comes directly from the Standard Normal Distribution, Z. In fact, any Normal Random Variable, X with Expected
value μ and Standard Deviation σ can be converted to a Standard Normal Distribution by using the formula:
Z = (X-μ)/σ
Example: Water Usage
Problem
The daily water usage per person in a town is normally distributed with a mean (expected value) of 20 gallons and a standard deviation of
5 gallons.

Part A
Determine the proportion of people who use between 15 and 25 gallons of water.

P(15 < X < 25) = P((15-20)/5 < Z < (25-20)/5) = P(-1 < Z < 1) = 0.6826

Part B
Determine the proportion of people who use between 10 and 30 gallons of water.

P(10 < X < 30) = P((10-20)/5 < Z < (30-20)/5) = P(-2 < Z < 2) = 0.9544

Part C
Between what two values would you expect to find about 95% of the water users?

Since P(-2 < Z < 2) = 0.9544, we can say about 95% of the water users are within two standard deviation of the mean, or that they use
between 10 and 30 gallons per day.
 The daily usageper person in a town is normally
More Water Usage Examples: distributed with a mean of 20 gallons and a
standard deviation of 5 gallons
Problem A 1
What is the probability that a person from the town
selected at random will use fewer than 18 gallons per
person per day? 2 Problem B

P(X < 18) = P(Z < (18-20)/5) = P(Z < -0.40) = 0.3446 What proportion of the people use between 18 and 24
gallons per person per day?

P(18 < X < 24) = P((18-20)/5 < Z < (24-20)/5) = P(-0.40 <
Problem C 3 Z < 0.80) = 0.4382
What percentage of the population uses more than 26.2
gallons per person per day?

P(X > 26.2) = P(Z > (26.2-20)/5) = P(Z > 1.24) = 0.1075 or 4 Problem D
10.75%
A special tax is going to be charged on the top 5% of
water users. Find the value of daily water usage that
generates the special tax.

X95 = 20 + 5(1.645) = 28.2 gallons per day

Problem A
Problem C

Problem B
Problem D
Additional Examples

Grading on the Curve

Professor Kurv has determined that the final
averages in his statistics course is normally
distributed with a mean of 77.1 and a standard
deviation of 11.2.

He decides to assign his grades for his current

course such that the top 15% of the students
receive an A.

The top 15% would be the finding the 85th

percentile. The corresponding Z value is 1.04.
The minimum grade for an A:
X = 77.1 + (1.04)(11.2) = 88.75 points.
Activity:
1. A researcher measures the time (in minutes) customers spend in a bank queue before being served. Is
this a discrete or continuous random variable? Justify your answer.
For the next questions, determine the probability distribution and answer the questions:
2. The time (in minutes) between arrivals of customers at a gas station follows an exponential distribution
with a mean of 4 minutes.

a) Determine the rate parameter λ.

b) Find the probability that the next customer will arrive in less than 3 minutes.

3. A bus arrives at a station every 30 minutes, and the waiting time for a randomly arriving passenger follows
a uniform distribution between 0 and 30 minutes.

a) Find the probability density function (PDF).

b) What is the probability that a passenger waits between 10 and 20 minutes?
4. The amount of tips the servers in an exclusive restaurant receive per shift is normally distributed with a
mean of $80 and a standard deviation of $10. Shelli feels she has provided poor service if her total tip for
the shift is less than $65. What percentage of the time will she feel like she provided poor service? Let y be
the amount of tips.

5. The heights of adult males in a city follow a normal distribution with a mean of 175 cm and a standard deviation of 6 cm.

What is the probability that a randomly selected male is taller than 180 cm?