Chapter 6

The Normal Distribution

David Chow
Oct 2021

1
 Learning Objectives

 In this chapter, you will learn:

 To compute probabilities from the normal distribution
 To determine whether a set of data is approximately normal

 The uniform distribution and the exponential

distribution are NOT covered

2
 Importance of a Normal Dist

 Many continuous variables

seem to be normally distributed
 Many discrete variables can be approximated by a normal
distribution
 Eg: The binomial distribution is symmetric under certain
conditions

 By the central limit theorem, sampling distributions are

approximately normal (more in ch7)

3
 Properties μ can take
any values
f(X) is density of X,
not probability of X

 Bell-shaped & symmetric

 By symmetry, μ = median = mode
 Location characterized by μ
 Spread characterized by σ
 Infinite range for the variable X, or

- < X < +
 In symbols, X ~ N (, 2)

4
 Probability and Area

 The bell-shaped curve is called a density function

 Probability of X is found by the corresponding area
under the density curve
 Hence, total area under the curve = 1
 Such area (probability) is found by checking
statistical tables

5
 Probability and Area
 Unlike discrete probability distributions, the probability of
a particular value from a continuous distribution is zero
 Eg: P (download time = 4s) = 0

 Why?
1. Probability is the area under the density curve
2. A continuous variable has infinitely many possible values

6
 Probability and Area
 If X is continuous, a probability is meaningful if it
corresponds to a range (or an interval) of X
 Eg: P ( download time < 4.0s)
“<“ or “”
 Eg: P (a  X  b), X = height
Does it matter?

7
 Shape

By varying μ and σ, we obtain

different normal distributions

Eg: Waiting time (X) at two university train stations:

1. At Shatin Univ, μ = 5 min, σ = 1.8 min
2. At Pokfulam Univ, μ = 4 min, σ = 1.5 min

Curve ___ gives the waiting time at Shatin Univ

Remark: Analogy of a straight line

8
 The Standardized Normal Z
 X is normally distributed, with mean=μ and SD=σ
 X can be transformed to the standardized normal (Z)
 Remember the transformation formula?

 Z is normally distributed, with mean=0 and SD=1

An Example on Transformation
• X = midterm score; µ = 80; σ = 10
• What happens if
1. The professor deducts 80 points from each student
2. All scores are multiplied by a constant (1/10 here)

9
 Eg: Waiting Time
 X = waiting time for customers at a bank
 X is normally distributed
 X ~ N (μ = 100, σ2 = 502)

1. Find the Z-value for X = 200s

2. What is X if Z = -1.5?

10
 Eg: Waiting Time (Cont)

• The transformation does not change the shape

• The same distribution can be expressed
• In original units (X), or in standardized units (Z)

11
 Eg: Height
 A Normal Distribution
 TOP: Suppose the height of adult
females is normally distributed,
 µ = 162.2cm
 σ = 6.8cm

 A standard question to ask

 RIGHT: What is the probability a
randomly selected female is
taller than 170.5cm?

 Question “re-phrased”
 BOTTOM: Prob (X < Z)

12
 Finding Normal Probability
Row: value of Z to the 1st decimal point
Column: value of Z to the 2nd decimal point

Standardized normal distribution Cumulative standardized normal

13
 Finding Normal Probability
Find the following areas if
the Z-values are known:

• A = Area in blue
• B = Area in white
• C = Column area
• D = Area in green

ANSWER
Suppose Z=1.2 or -1.2

• A = P(Z<1.2) … read from the Z-table = 0.8849

• B = 1 - Area in A
• C = P (0<Z<1.2) = P(Z<1.2) – P(Z<0)
• D: Area B gives the upper-tail area
By symmetry, Area D = Area B

14
Eg1: P (Z < 2)

P (Z < 2.00) = 0.9772

Q: So the Empirical Rule is wrong?

15
Eg2: Standardized Normal Distribution

a. Find the standard deviation of the normally distributed variable x

b. What are the required probabilities?

ANSWER

16
 Eg3: Downloading Time

 X = downloading time (in s) of an apps

X ~ N (μ = 8.0, σ2 = 5.02)
 Find P(X < 8.6)

μ=0
σ=1

Next, find P(8.0 < X < 8.6)

17
 Eg3: Downloading Time
(Find X Given the Probability)

 Find X such that 20% of download

times are less than X

 First, use the table to find the Z-value of

the given probability of 0.20. Z = ____

 Second, convert the Z-value to X units

using the transformation formula

 So 20% of the download times are ____

18
 Finding Normal Probability
 The Cumulative Z-Table gives the probability of ____

 To find P (a < X < b) where X ~ N (μ, σ2)

 Translate X-values to Z-values
 Check the required probability from the table
 A visual check is often useful

19
 Assessing Normality
 There are different ways to assess normality

1. Visual check by a histogram or a box-and-whisker plot

2. Check the descriptive measures:

 Similar values of mean, median and mode
 Range approximately equals 6σ

20
 Assessing Normality

3. Use the Empirical Rule:

 About 67% of the observations lie within μ ± σ
 About 95% of the observations lie within μ ± 2σ

4. Use the normal probability plot

21
 Interpreting Normal Probability Plot

A normal probability plot is approximately

linear for a normal distribution

A normal probability plot from a skewed distribution will be non-linear

 Review Questions: T or F
True or False

1. In a standard normal distribution, the probability that Z is

greater than 0.5 is 0.5

2. In a standard normal distribution, the probability that Z is

greater than 1.96 is 2

3. For a continuous random variable x, the probability

density function f(x) represents the probability at a given
value of x

23
 Appendix
Cumulative-Z
& Excel Commands

24
 Excel Command
 FIND AREA  FIND X- or Z- VALUES
 Find cumulative probability (i.e., area  Find X- or Z- values given the
from the left) given X- or Z- values cumulative probability (i.e., area
 =NORMDIST(X, , , true) returns from the left)
the cumulative probability of a normal  =NORMINV(cumulative
distribution
probability, , )
 Eg: =NORMDIST(5, 4, 1, true)
gives P(X < 5, given  = 4,  = 1),  =NORMSINV(cumulative
I.e., 0.8413
probability)
 =NORMSDIST(Z) gives the  Eg: =NORMSINV(0.5)
cumulative probability of a
standardized normal.
 Eg: =NORMSDIST(0.12) gives
Re-do Eg3 on p.18
P(Z < 0.12)

25
26