NORMAL

DISTRIBUTION
Understanding normal distribution
What is a distribution?
• Imagine we have some data. When we plot these data, we can see
some patterns emerging. These patterns are what we called as
‘data distribution’, because we can see how our data are
‘distributed’ across a range of values.
• In other words, A distribution is simply a collection of data.
• You may find that a lot of data were clustered around the ‘middle’
values. On the other hand, at the edges of the distribution we can
expect fewer observations (Where we can find very high or very low
values).
Some of the •Binomial distribution
discrete •Poisson distribution
distributions
Binomial Distribution Parameters
probability of success: p ∈ ℝ: 0 ≤ p ≤ 1
number of trials: n ∈ ℤ: n ≥ 0

0.25 Support
x ∈ {0, 1, ..., n}
(X = x) 0.2
Example Moments
probability mass E(X) = np
0.15 Var(X) = np(1-p)
plot for
binomial
Probability

distribution 0.1
when free
parameter 0.05
p=0.5, fixed
parameter=10,
range 0-10 0

0 2 4 6 8 10

X (example: number of successes)

Poisson Distribution
“Poisson” means
“fish” in French.

Parameters
0.4
rate: λ ∈ ℝ: λ > 0

Support
(X = x)
0.3
Example x ∈ {0, 1, 2, ...}
probability mass
plot for Poisson Moments
0.2
Probability

distribution E(X) = λ
when λ=1 , Var(X) = λ
range 0 to 5 0.1

0
0 1 2 3 4 5

x
 • Normal distribution
Some of the • T-distribution
continuous • F-distribution
distributions • Chi-square distribution
Normal Distribution Parameters
mean: μ ∈ ℝ
variance: σ2 ∈ ℝ+

Support
0.4 x∈ℝ
When =0,
Moments =1, the normal
0.3 E(X) = μ distribution is
Example of Var(X) = σ2 called the
normal
Density

“Standard
distribution 0.2 Normal
when =0, =1,
Distribution”
range: -3 to 3
0.1

-3 -2 -1 0 1 2 3

x
 Parameters

T- distribution
degree of freedom: df ∈
ℝ+
non-centrality: ncp ∈ ℝ

0.3 Support
x∈ℝ

Moments
Density

Example of 0.2 E(X) = does not exist unless df > 1

t Var(X) = does not exist unless df > 2
distribution
when df=1,
0.1 T distribution is also
ncp=o,
bell-shaped. Other
range: -3 to bell-shaped
3 distribution includes :
Cauchy, logistic. When
0 t distribution having
df=1, it became a
-3 -2 -1 0 1 2 3 Cauchy (Lorentzian)
distribution.
x
What is the degree of freedom (df)?
Lets imagine you have 7 different shirts. You want to wear a shirt of different colour everyday for the week.

On the 1st day, you can pick any of the 7 shirts. On the 2nd day, you can choose from the 6 remaining
shirts (since you don’t want to wear the same shirt). On the 3 rd day you can choose from 5 shirts, and so
on. On the 6th day you still have the freedom to choose which remaining 2 shirts you would wear.

However on the last day (7th day), you have no choice but to wear the one remaining shirt.

Think of the “degree of freedom” as how many times you have the “freedom” to choose. Only on the
7th day you don’t have the freedom to choose your shirt.

So, in a single sample like this with n=7, your df=6, which is n-1.
Now, your friend noticed that you wear different shirts everyday and he be like “ Man, That’s so cool!
I want to do that too!” And so he went to buy 7 different shirts.
You said “ Yeah man lets wear a different shirt every day for next week.

You

Your friend

From the 1st day till the 6th day, both of you have the freedom to choose your shirt. However on the last
day (7th day), again, you have no choice but to wear the one remaining shirt. Same goes for your friend.

Both of you can choose your shirt for 6 days out of 7 days. The degree of freedom is simply the number
of days you both can choose a shirt to wear.

So, in two samples like this with (total number of shirts = 14), the degree of freedom is (n₁ -1) + (n₂-1) =
(7-1) + (7-1)= 12.
Application of Degree of Freedom (df)
• Degrees of freedom are important in inferential statistics (we estimate or make
inferences about the population based on sample).
• Typically the degrees of freedom is related to the size of the sample. Because a
higher degrees of freedom implies you have a larger sample size.
• Having a high degree of freedom means more power to reject the false null
hypothesis (having a significant result
• You will also need the df for the calculation of t-ratio
• Different statistical tests can have a different formula for df.
F-distribution
Parameters
degree of freedom: df 1 ∈ ℝ+
degree of freedom: df 2 ∈ ℝ+
0.6
non-centrality: ncp ∈ ℝ
0.5
Support
x ∈ ℝ+
0.4
Example of t
Density

distribution Moments
0.3
when df1=4, not available
df2=2 ncp=o,
0.2
range: 0 to 3
0.1

0.0 0.5 1.0 1.5 2.0 2.5 3.0

x
Chi-square distribution
Parameters
0.2 degree of freedom: k ∈ ℝ+
non-centrality: λ ∈ ℝ
0.15 Support
Example of chi- x ∈ ℝ+
Density

square
distribution 0.1 Moments
when k=5, λ=0, E(X) = k + λ
range: 0 to 15 Var(X) = 2(k + 2 λ)
0.05

0 5 10 15

x
 NORMAL
DISTRIBUTION
Why it is called “normal” distribution
• Over many years, statisticians noticed that data (either from samples or
populations) keep producing similar patterns.
• The Normal distribution been studied for hundreds of years with many different
names. Charles S. Peirce (in 1873) and Francis Galton (in 1877) were among the
first mathematicians to use the term “normal”.
• Normal distribution is also called Gaussian distribution, named after Carl Gauss
(a German mathematician) who first described it.
• Mathematicians have developed standard statistical distributions that describe
these frequently emerging patterns. Standard distributions are important in
statistical testing because researchers use them as reference distributions. In
other words, we can use them to compare data/samples.
Normal distribution
Perfectly normal distribution:
0.4 Mean = Median = Mode

0.3
Symmetric sides
Density

0.2 A Standard Normal

Distribution:
A normal distribution
0.1 with mean = o and
SD=1
tail
0

-3 -2 -1 0 1 2 3

x
 Distributions with different parameters
μ=0 μ = -2 μ = 1.5
σ=1 σ = 0.5 σ=2
0.4 0.8 0.2

0.3 0.6 0.15

Density

Density
0.2 Density
0.4 0.1

0.1 0.2 0.05

0 0 0
-4 -2 0 2 4 -2 0 2 -4 -2 0 2 4 6
-4 4
x x x
 Symmetric around the mean.

Perfectly normal distribution: The mean = median = mode.

The area under the curve = 1.0.

8 Characteristics Denser in the centre, less dense at the sides (tails).

of normal
distributions It has two parameters: the mean (μ) and the standard deviation (σ).

68% of the area is within one standard deviation from the mean.

Empirical Rules 95% of the area is within two standard deviation from the mean.

99.7% of the area is within three standard deviation from the mean.
 68% of values are within one standard deviation (SD) from both side of the
mean (also written as ±1 SD):

0.4 A normal
distribution with
mean = o and
0.3 SD=1
Density

0.2
Mean
0.68
0.1

-3 -2 -1 0 1 2 3

Standard Deviation (SD)

Empirical Rules
 95% of values are within two standard deviation (SD) from both side of the
mean (also written as ±2 SD):

A normal
0.4 distribution with
mean = o and
SD=1
0.3
Density

0.2 Mean

0.95
0.1

-3 -2 -1 0 1 2 3
Empirical Rules x
 99.7% of values are within three standard deviation (SD) from both side of
the mean (also written as ±2 SD):

0.4 A normal
distribution with
mean = o and
0.3 SD=1
Density

0.2 Mean

0.997
0.1

-3 -2 -1 0 1 2 3

Empirical Rules x
Example: Empirical Rule
0.4
Assume that the
mean weight of
0.3
medium size 1 1

Density
dogs in a dog
shelter is 0.2
normally
distributed with 0.1
2
2
a mean weight
3 3
of 9.5kg and 0
standard 9.5
deviation of 1.1. Weight (kg)
Complete the
graph
 Solution: Empirical Rule
0.4

0.3
1 1
Density

0.2 Here is 3 SD
more, Therefore
0.1
2 2 9.5 + 3(1.1) = 12.8
3 3
0
9.5
6.2 7.3 8.4 10.6 11.7 12.8
Weight (kg)
Here is 1 SD less,
Therefore 9.5 - 1.1 = 8.4
Measures of shape (Skewness & Kurtosis)
• We use central tendencies and measures of dispersion to describe the distribution,
but they are not sufficient to describe how your distribution looks like (as
compared to a standard normal distribution). Is your graph too tall? Is the peak
located at the center like a normal distribution?
• For this purpose we use Skewness and Kurtosis as the measures of the shape of
the distribution.
• Skewness assesses whether the distribution is symmetrical. If the distribution
looks like it is “leaning” toward the right or left tail of the distribution, then we can
say that this distribution is skewed.
• Kurtosis tells you how tall and sharp the central peak is, as compared to a
standard normal distribution
Skewness
Skewness means lack of symmetry. If
you cut the graph at the mean (red dash
Mode Mode
line), do you have equal portion (left vs
right)?
Median Median
Tips: If you are
Mean confused which one Mean
is + or – Skewed,
look at the tail

+ -

-4 -2 0 2 4 -4 -2 0 2 4
x x

Positively-skewed Distribution Negatively-skewed Distribution

 Kurtosis Positive Kurtosis
Note:
Kurtosis: Measure of “peakness”
(Leptokurtic) Excess kurtosis = Kurtosis - 3

Leptokurtic: kurtosis >3 (excess kurtosis

>0). Central peak is higher and sharper.
Normal Distribution
(Mesokurtic)
A normal distribution has
kurtosis = 3 (excess kurtosis =
Platykurtic: Kurtosis 0). Mesokurtic : kurtosis ≈3
<3 (excess kurtosis
Negative Kurtosis
(excess ≈0).
<0). Central peak is (Platykurtic)
lower and broader.

-4 -2 0 2 4

x
 • Central Limit Theorem stated that – It
Central Limit doesn’t matter how your distribution
initially looks like; the distribution of the
Theorem sample means will approximate normal
(CLT) distribution as your sample size is
getting larger.
 Central Limit Theorem
• Example: This distribution is not a
normal distribution.
• Suppose we randomly take several
samples from this list [1,2,3,4,5,6] ,and
each sample contain 4 values and then
Density

calculate the sample mean:

• Sample 1 = [1,1,3,6] mean 1 = 2.75
• Sample 2 = [ 3,4,3,1] mean 2 = 2.75
• Sample 3 = [1,1,6,6] mean 3 = 3.5
1 2 3 4 5 6
value
 Central Limit Theorem
• Sample 1 = [1,1,3,6] mean 1 = 2.75
• Sample 2 = [3,4,3,1] mean 2 = 2.75
Distribution of sample means
• Sample 3 = [1,1,6,6] mean 3 = 3.5
2.5

2 • We plot these sample mean in a

frequency distribution .
1.5
Density

• If we keep doing this over and over,

1 eventually your distribution will
0.5
approximate a normal distribution.
• Demonstration: CLT Simulator
0
2.75 3.5
Sample means
http://onlinestatbook.com/stat_sim/sam
pling_dist/index.html
Now that you know about normal
distribution. What’s next?
Normality testing
• Some statistical tests, such as t-test and correlation requires your data to be
normally distributed (or nearly normal).
• There are many methods to assess whether your data follow a normal
distribution.
• Some of the common methods to do this is with Kolmogorov–Smirnov test or
the Shapiro–Wilk test, Q-Q plot, or looking at the skewness and kurtosis.
• A normal distribution has skewness and excess kurtosis of zero (0), so if your
distribution is close to those values then it is probably close to normal (not
more than +2 or less than -2).
Now that you know about normal
distribution. What’s next?
Calculate Probability
You can also use normal distribution to calculate Probability (P) related to
continuous variables. To determine the Probability of an event, we look at the
“area under the curve” . We have seen this earlier. Remember the 68-95-99 %
graph?
Compare data points from different distribution
To compare data points from different distribution, Data Standardization, in
statistics is simply means putting different variables on the same scale. It is easier
to interpret these variables when they are in the same scale. To standardize a
distribution, we apply a formula (i.e. z-score formula) to all observations so that
the SD= 1.
 Z-score
• A z-score describes the position of a raw score/value in terms of its distance
from the mean, when measured in standard deviation units.
• The z-score is positive if the value (x) is higher than the mean, and negative if
the value (x) is lower than the mean.
Example: Z-score of 1.5 means the raw score/value is 1.5 standard deviation
above/higher than the mean. Z-score of -1.5 means the raw score/value is 1.5
standard deviation below/lower than the mean.
• Even if your data are not in the same scale, you can convert them into the same
scale using z-scores (how many SD from the mean). The resulting z-scored
versions of data will all have SD = 1.
Example: computing z-score
• Grades on a statistics exam are normally distributed with μ = 81 and =6.3.
Calculate z-score for each of the following grades.
a) 65 μ = 81
0.06
b) 83
c) 93 Density 0.04

d) 100
0.02

𝑥 −𝜇
𝑧= 0
𝜎 60 70 80 90 100
Exam Score
 Solution
a)
Mean=81
b)
0.06
c)

d)
Density

0.04

0.02

For score 100, z score is

0
3.02. Therefore 100 mark
60 70 80 90 100 is 3.02 SD more than the
Exam Score mean (81).
Example: comparing z-scores
• Harry took 2 tests. He scored 172 in Defense Against the Dark Art exam and
37 in Potion exam. Which exam he did relatively better?

Exam Mean mark for Standard Deviation

the class , μ
Defense Against the Dark 151 10
Art
Potion 25.1 6.4
Solution: comparing z-scores
Exam Mean mark for Harry’s Mark, x Standard
the class , μ Deviation, 
Defense Against the Dark Art 151 172 10
Potion 25.1 37 6.4

Defense Against the Dark Art Potion

𝑥 −𝜇 172 −151 𝑥 −𝜇 25.1 −37

𝑧= = =2.1 𝑧= = =1.86
𝜎 10 𝜎 6.4
His mark for Defense Against the Dark Art exam is 2.1 SD above the mean.
On the other hand, his Potion mark is 1.86 SD above the mean. Therefore, he
did relatively better in Defense Against the Dark Art exam
 DETERMINE THE AREA
UNDER THE CURVE USING
Z-SCORE AND Z-TABLE
Example: Area under the curve
• The heights for a portion of first years Hogwarts student are normally
distributed with mean=150 and SD=20cm. Ron is 161.4cm.
a) What is the proportion of students shorter than Ron?
b) What is the proportion of students taller than Ron?
c) Malfoy’s height is 125cm. What proportion of students are between
Malfoy’s and Ron’s height?

Tips: Lets use z-table for this. You can download z-table from the website
donaldstephen.wixsite.com/statistics
 What is the proportion of students shorter than Ron?
Situation 1: Proportion LOWER than x (from - to x)

Step 1: Get a z-score for x. If you

0.02 do not already have a z-score, use
We are trying to to calculate it.
0.015 find the area
under the curve
for < 161.4
Density

0.01 (shaded area)

0.005

100 120 140 150 161.4 180 200

Height (cm)
What is the proportion of students shorter than Ron?
Situation 1: Proportion LOWER than x (from - to x)

Step 2: Look up value of z in the z table. For example, the proportion for z= 0.57 is
highlighted in the z-table. When z= 0.57, the area under the curve is 0.7157

Z 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239 0.5279 0.5319 0.5359
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636 0.5675 0.5714 0.5753
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026 0.6064 0.6103 0.6141
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406 0.6443 0.6480 0.6517
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772 0.6808 0.6844 0.6879
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224
0.6 0.7257 0.7291 0.7324 0.7357 0.7389 0.7422 0.7454 0.7486 0.7517 0.7549
0.7 0.7580 0.7611 0.7642 0.7673 0.7704 0.7734 0.7764 0.7794 0.7823 0.7852
0.8 0.7881 0.7910 0.7939 0.7967 0.7995 0.8023 0.8051 0.8078 0.8106 0.8133

71.57 % of students are shorter than Ron

What is the proportion of students taller than Ron (161.4) ?
Situation 2: Proportion HIGHER than x (from x to )

Step 1: Get a z-score for x. If you

We are trying to do not already have a z-score, use
find the area to calculate it.
under the curve
for > 161.4
(shaded area)

150 161.4
What is the proportion of students shorter than Ron?
Situation 2: Proportion HIGHER than x (from x to )

Step 2: Look up value of z in the z table. The same way how we get our z-score in
previous example.

Z 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239 0.5279 0.5319 0.5359
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636 0.5675 0.5714 0.5753
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026 0.6064 0.6103 0.6141
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406 0.6443 0.6480 0.6517
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772 0.6808 0.6844 0.6879
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224
0.6 0.7257 0.7291 0.7324 0.7357 0.7389 0.7422 0.7454 0.7486 0.7517 0.7549
0.7 0.7580 0.7611 0.7642 0.7673 0.7704 0.7734 0.7764 0.7794 0.7823 0.7852
0.8 0.7881 0.7910 0.7939 0.7967 0.7995 0.8023 0.8051 0.8078 0.8106 0.8133
 What is the proportion of students taller than Ron (161.4) ?
Situation 2: Proportion HIGHER than x (from x to )

0.7157 Step 3: Note that the area

obtained is from - to x.
-
Therefore, this Since we know that the
area = 1- 0.7157 = whole area under the
0.2843 curve =1; to get the area
from x to  , we just need
to subtract 0.7157 area
from 1.

150 161.4

- Whole area = 1 
What is the proportion of students between Malfoy’s (125) and Ron’s (161.4) height?
Situation 3: Proportion between 2 values (from x1 to x2)

Step 1: Get a z-score for x1 and

We are trying to find
x2. If you do not already have a z-
the area under the
score, use to calculate it.
curve from 125 to
161.4 (shaded area)

150
X1 = 125 X2 = 161.4
What is the proportion of students shorter than Ron?
Situation 3: Proportion between 2 values (from x1 to x2)

Step 2: Look up value of z1 and z2 in the z table. The same way how we get our z-score in
previous example. Since z1 is negative z score, look up for it on negative z-table

When z = , area is 0.1056

Z 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

-1.7 0.0446 0.0436 0.0427 0.0418 0.0409 0.0401 0.0392 0.0384 0.0375 0.0367
-1.6 0.0548 0.0537 0.0526 0.0516 0.0505 0.0495 0.0485 0.0475 0.0465 0.0455
-1.5 0.0668 0.0655 0.0643 0.0630 0.0618 0.0606 0.0594 0.0582 0.0571 0.0559
-1.4 0.0808 0.0793 0.0778 0.0764 0.0749 0.0735 0.0721 0.0708 0.0694 0.0681
-1.3 0.0968 0.0951 0.0934 0.0918 0.0901 0.0885 0.0869 0.0853 0.0838 0.0823
-1.2 0.1151 0.1131 0.1112 0.1093 0.1075 0.1056 0.1038 0.1020 0.1003 0.0985
What is the proportion of students shorter than Ron?
Situation 3: Proportion between 2 values (from x1 to x2)

Step 2: Look up value of z1 and z2 in the z table. The same way how we get our z-
score in previous example.

Z 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239 0.5279 0.5319 0.5359
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636 0.5675 0.5714 0.5753
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026 0.6064 0.6103 0.6141
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406 0.6443 0.6480 0.6517
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772 0.6808 0.6844 0.6879
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224
0.6 0.7257 0.7291 0.7324 0.7357 0.7389 0.7422 0.7454 0.7486 0.7517 0.7549
0.7 0.7580 0.7611 0.7642 0.7673 0.7704 0.7734 0.7764 0.7794 0.7823 0.7852
0.8 0.7881 0.7910 0.7939 0.7967 0.7995 0.8023 0.8051 0.8078 0.8106 0.8133
What is the proportion of students between Malfoy’s (125) and Ron’s (161.4) height?
Situation 3: Proportion between 2 values (from x1 to x2)
0.7157

-

Step 3: To get area between x1 and

x2: larger area - smaller area

0.1056 Shaded area:

0.7157 - 0.1056 = 0.6101
-
Reminder: 0.6101 is also 61.01%

X1 = 125 150
X2 = 161.4
 SOLVING OTHER
FINDING Z-SCORE FOR A PERCENTILE PROBLEMS
THRESHOLD FOR PERCENTILE INVOLVING Z
SCORES:
Example: Finding Z score for a percentile
The distribution of resting pulse rates of all students at UNIMAS was
approximately normal with mean=80bpm and the standard deviation of
9bpm.
UNIMAS nurse plans to do health screening for students whose pulse rate is in
the top 30%.
What is the minimum resting pulse rate for UNIMAS students who will
receive health screening? Round the answer to the nearest whole number.
 Solution: Finding Z score for a percentile
70%
0.05
30%
We need to find the
0.04 z-score corresponds
When pulse rate is in the to probability 0.7.
top 30%, will do screening.
Density

0.03
So from z-table, find
0.02 probability nearest
to 0.7 (a little bit
0.01 more is ok).

0
50 60 70 80 T 90 100 110
x
Mean=80 , SD=9. We want to find the minimum resting pulse rate that will receive
health screening
Solution: Finding Z score for a percentile
We need to find the z-score corresponds to probability 0.7.
So from z-table, find probability nearest to 0.7 (a little bit more is ok).

Z 0 0.01 0.02 0.03 0.04 0.05 0.06

0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123
0.6 0.7257 0.7291 0.7324 0.7357 0.7389 0.7422 0.7454
0.7 0.7580 0.7611 0.7642 0.7673 0.7704 0.7734 0.7764

0.7019 is the closest to 0.7. The z-score corresponds to probability 0.7019 is 0.53.
Now the question is – What value is 0.53 SD from the mean?
 Solution: Finding Z score for a percentile
70%
0.05
30%
0.04
We know that this point here, T is
Density

0.03 0.53 SD more than the mean

0.02
Threshold value, T = Mean + zscore*SD
0.01 80 + 0.53(9) = 84.77  85 beats per minute

0
50 60 70 80 T 90 100 110
x
Mean=80 , SD=9. The min resting pulse rate that will receive health screening = 85
Example: Threshold for low percentile
• The distribution of average wait time at the UNIMAS front gate
security at 8.00am is approximately normal with mean=185
seconds and standard deviation of 11 seconds.
• Sheldon only likes to use the front gate where average time is in
the bottom 10%.
• What is the maximum average waiting time where Sheldon likes to
drive through the security at front gate?
 Solution: Threshold for low percentile
0.04 Mean=185, SD = 11.

So, when we ask “What is the

0.03 When waiting maximum average waiting time
time is in this where Sheldon likes to drive through
region (bottom
Density

the security at front gate?”, we are

0.02 10%), Sheldon basically finding this point.
would pass
through.
0.01 First we need to find what
z-score corresponds with
10%, which is also 0.1.
0 Since x is below mean, we
use negative z-table.
150 160 170 180 190 200 210 220

x
Solution: Threshold for low percentile
We need to find the z-score corresponds to probability 0.1 from negative z-table.
So from z-table, find probability nearest to 0.1

Z 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

< = -3.9 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000

-1.3 0.0968 0.0951 0.0934 0.0918 0.0901 0.0885 0.0869 0.0853 0.0838 0.0823
-1.2 0.1151 0.1131 0.1112 0.1093 0.1075 0.1056 0.1038 0.1020 0.1003 0.0985
-1.1 0.1357 0.1335 0.1314 0.1292 0.1271 0.1251 0.1230 0.1210 0.1190 0.1170
-1 0.1587 0.1562 0.1539 0.1515 0.1492 0.1469 0.1446 0.1423 0.1401 0.1379
-0.9 0.1841 0.1814 0.1788 0.1762 0.1736 0.1711 0.1685 0.1660 0.1635 0.1611

Here we can see that 0.1003 is the closest to 0.10. but we don’t want this value because we want
the bottom 10% (less than 0.10). Therefore, we should take 0.0985. Hence, z-score= -1.29
 Solution: Threshold for low percentile
0.04 Mean=185, SD = 11.
We know that z-score is -1.29. so,
we know that the threshold is
Maximum
0.03 1.29 SD BELOW the mean.
average waiting
time where
Density

Sheldon likes to X = 185 – 1.29(SD)

0.02 = 185 -1.29(11)
drive through
the security = = 170.81
170 s Note: If the question is
0.01 asking you to round the
result into the closest
whole number, then the
0 answer is 170, because we
150 160 170 180 190 200 210 220 don’t want to go over that
threshold.
x