Part VIII.

Statistical Process
Control
e. Pareto Chart
f. f. Histogram

Zaldy R. Mendoza Jr.

 Histograms

Summary of column graphs so far

A Bar Chart A Frequency Diagram
Time taken to complete a Table Test
Sales of Paddling Pools per quarter
100
6
80 5

Frequency
60 4
40 3
2
20
1
0
1st Qtr 2nd Qtr 3rd Qtr 4th Qtr 0 2 4 6 8 10
Time in Minutes
Spot the differences from a bar chart to a frequency diagram:
• There are no gaps between the bars
• The horizontal axis has a continuous scale as it represents
continuous data
• The width of the bars is the full width of the class interval
• The class intervals (the width of the bars) is the same for all bars
 Histograms
• A histogram is a bar graph of a distribution of
variables
• Each bar represents an attribute or characteristic
of a problem or situation, and the height of the bar
represents its frequency
 Histograms

A histogram is a frequency diagram with unequal class intervals.

The area of the bar represents the frequency
Why have different class intervals?
Sometimes data will be collected where there may be very little
data in a class interval, and yet in another class interval of the
same width.
e.g. a survey was conducted to find the ages of surfers
 Histograms

e.g. a survey was conducted to find the ages of 60 surfers, the results
are in the table below:
Age Frequency
 0 < x ≤ 15 4 
 15 < x ≤ 20  14
 20 < x ≤ 25  12
 25 < x ≤ 30  8
 30 < x ≤ 40  12
40 < x ≤ 70  10

Notice how the class intervals are different

If you kept the class interval of 5 (as for ages 15 to 30) you may
find several empty class intervals between 40 and 70
because the class width is different if we used frequency for the
y-axis the sizes of the bars would represent the frequencies
disproportionately.
 Histograms

We use frequency density to show the height.

The frequency is the area of the bar
The width of the bar is the class width
Area of a rectangle = length × width
Frequency density

Frequency frequency = frequency density × class width

=
AREA rearranging this equation

frequency density = frequency

Class width
class width
 Histograms
Draw a histogram to represent this data
Age Frequency Class Width Frequency Density
1
 0 < x ≤ 15 5  15  3 
 15 < x ≤ 20  15 5  3 
 20 < x ≤ 25  10 5  2 
 25 < x ≤ 30  8 5   
1.6
 30 < x ≤ 40  12 10  1.2
 
1
40 < x ≤ 70  10 30  3 

3
Frequency density

0 10 20 30 40 50 60 70
Age (years)
 Histograms
From this histogram you could be asked further questions:
Estimate how many people between the ages of 35 and 55 surf.
3

Identify the area of the

Frequency density

2 bars to which this applies.

Remember:
1
frequency is the area

0 10 20 30 40 50 60 70 1.2
Age (years)
1
35 55 3

5 15
Find the area separately of each bar
5 × 1.2 = 6 15 × 1 = = 5 Estimate of the frequency = 6 + 5 = 11
3
 Histograms
Draw Histograms for these tables:
Temperature
Frequency
Class Frequency 10
(nearest oC) Interval Density
8

Frequency Density
10 ≤ t < 11 5   1   5
11 ≤ t < 12 7   1   7 6
12 ≤ t < 15 18   3   6
4
15 ≤ t < 16 4   1   4
16 ≤ t < 17 3   1   3 2
17 ≤ t < 20 6   3   2 0
10 11 12 13 14 15 16 17 18 19 20
Temperature oC

Class Frequency
50
Height (cm) Frequency
Interval Density 40
Frequency Density

151 ≤ h < 153 64 2 32

153 ≤ h < 154 43
30
1 43
154 ≤ h < 155 47 1 47 20
155 ≤ h < 159 96 4 24
10
159 ≤ h < 160 12 1 12
0
151 152 153 154 155 156 157 158 159 160
Height (cm)
 Histograms
Draw Histograms for these tables:
25
Time
Class Frequency
(nearest Frequency 20

Frequency Density
Interval Density
minute)
0≤t<8 72   8   9
15
9 ≤ t < 12 84   4   21 10
12 ≤ t < 15 54   3   18
5
15 ≤ t < 20 35   5   7
0
0 2 4 6 8 10 12 14 16 18 20
Time (mins)

25
Age (nearest Class Frequency 20
Frequency
year) Interval Density Frequency Density

11 ≤ a < 14 51   3   17 15
14 ≤ a < 16 36   2   18
10
16 ≤ a < 17 12   1   12
17 ≤ a < 20 20   3   6.7 5
0
11 12 13 14 15 16 17 18 19 20
Age (years)
 Histograms
Draw Histograms for these tables:
Time
Class Frequency
(nearest Frequency
Interval Density
minute)
0≤t<8 72    
9 ≤ t < 12 84    
12 ≤ t < 15 54    
15 ≤ t < 20 35    

Age (nearest Class Frequency

Frequency
year) Interval Density
11 ≤ a < 14 51    
14 ≤ a < 16 36    
16 ≤ a < 17 12    
17 ≤ a < 20 20    
 Pareto Charts
• A Pareto chart is a histogram that can help you
identify and prioritize problem areas

• Pareto analysis is also called the 80-20 rule,

meaning that 80 percent of problems are often due to
20 percent of the causes
Figure 8-7. Sample Pareto Chart
 What Is Pareto Analysis?
• Pareto analysis is a technique used for business
decision-making, but which also has applications
in several different fields from welfare
economics to quality control. It is based largely
on the "80-20 rule." As a decision-making
technique, Pareto analysis statistically separates a
limited number of input factors—either desirable
or undesirable—which have the greatest impact
on an outcome.
• Pareto analysis is premised on the idea that
80% of a project's benefit can be achieved by
doing 20% of the work—or, conversely, 80%
of problems can be traced to 20% of the
causes. Pareto analysis is a powerful quality
and decision-making tool. In the most
general sense, it is a technique for getting the
necessary facts needed for setting priorities.
• As a decision-making technique, Pareto analysis
statistically separates a limited number of input
factors—either desirable or undesirable—which
have the greatest impact on an outcome.
• Pareto analysis states that 80% of a project's
benefit or results are achieved from 20% of the
work—or conversely, 80% of problems can be
traced to 20% of the causes.
• With Pareto analysis, each problem or benefit is
given a numerical score based on the level of
impact on the company; the higher the score, the
greater its impact.
• Modern-day applications of Pareto analysis are
used to determine which issues cause the most
problems within various different departments,
organizations, or sectors of a business.
• By allocating resources to issues with higher
scores, companies can use Pareto analysis to
solve problems more efficiently because they
can target those problems that have a greater
impact on the business.
 Understanding Pareto Analysis
• In 1906, Italian economist Vilfredo Pareto
discovered that 80% of the land in Italy was owned
by just 20% of the people in the country. He
extended his research and determined that this
disproportionate wealth distribution was the same
across all of Europe. The 80-20 rule was formally
defined as follows: The top 20% of a country’s
population accounts for an estimated 80% of the
country’s wealth or total income.
• Joseph Juran, a Romanian-American
business theorist, discovered Pareto’s
research in 1937, approximately 40 years
after it was published. Juran proceeded to
rename the 80-20 rule as "Pareto’s Principle
of Unequal Distribution."
• Juran extended Pareto’s principle to the business
world in order to understand whether the rule could be
applied to problems faced by businesses. He observed
that in quality control departments, most production
defects resulted from a small percentage of the causes
of all defects. So, by extension, 80% of the problems
are caused by 20% of the defects; the implications of
Juran's work are that if you focus on fixing that 20%,
you could have a big impact with minimal effort
Here are some scenarios relevant
to businesses where Pareto
analysis might be applicable:
• Sharing information about defects/errors
with high priority stakeholders
• Prioritizing defects or tasks according to
their severity, i.e. according to their impact
on a system or business
• Analyzing data or errors/defects
 Steps of Pareto Analysis
• Identify the problem or problems
• List or identify the cause of the issues or
problems, noting that there could be multiple
causes
• Score the problems by assigning a number to
each one that prioritizes the problem based
on the level of negative impact on the
company
• Organize the problems into groups, such as
customer service or system issues
• Develop and implement an action plan,
focusing on the higher scored problems
first, in order to solve the problems
• Not all problems will have a high score,
and some smaller problems may not be
worth pursuing initially. By allocating
resources to high-impact issues or higher
scores, companies can solve problems
more efficiently by targeting the issues
that have a major impact on profits, sales,
or customers.
• Pareto analysis shows that a
disproportionate improvement can be
achieved by ranking various causes of a
problem and concentrating on the
solutions with the largest impact.
 Advantages and Disadvantages
of Pareto Analysis
• In the most general sense, the advantage of
Pareto analysis is that it helps to identify
and determine the root causes of defects or
problems. Because of this, businesses are
able to eliminate or resolve defects or errors
with the highest priority first.
• Pareto charts can specifically help determine the
cumulative impact of a problem. Cumulative impact
results from effects caused by a problem happening over
a long period of time. Pareto charts are especially useful
for businesses or organizations because they can use
them to plan the measures or actions that need to be taken
in order to amend the problems. For this reason, Pareto
charts can sharpen problem-solving and decision-making
skills: problems related to a defect or error can be
distilled into cohesive facts.
• The main disadvantage of Pareto analysis is
that it does not provide solutions to issues; it
is only helpful for determining or identifying
the root causes of a problem(s). In addition,
Pareto analysis only focuses on past data.
While information about past errors or
problems is useful, it's not a guarantee that
it will be relevant in future scenarios.
• A final disadvantage of Pareto charts is that
they can only show qualitative data that can
be observed; they cannot be used to
represent quantitative data. For example,
Pareto charts cannot be used to calculate
the mean, the standard deviation, or the
average of the data, its variability, or
changes in the measured attribute over time.
Bar graph