Explanation 9

(1) What is a Pareto Diagram?

 Example 
1) A Pareto Diagram is a bar chart constructed as follows:
Plug
- Segment the range of the data into groups (also called
segments, bins, categories) based on the type of
Part No.: BPR5S
occurrences or causes (e.g., such workplace problems
Period: 06/04~08/31
as nonconforming parts, defects, complains, accidents,
Inpector: Takeshita
etc.).
n = 620
- Arrange them from the most frequent to the least
frequent (e.g., nonconformance counts, amount of
cumulative loss for each group). (%)
2) A Pareto Diagram shows that, while there are many (Pieces) 100
600
causes for nonconformance or loss, it is only few items
that account for the most cases. It visually depicts which 500

items are more significant. Defective unit

400
3) You’ll know where to focus your efforts -- and
300 50
you’ll be in the fast lane to the problem-solving.
200

100

0 0
Burrs
Plating

Chatter marks
Damage

Contamination

(2) What a Pareto Diagram can tell you about

1) A Pareto Diagram depicts which situations are more significant.

2) A Pareto Diagram makes you see the magnitude of problem in descending order with regard to each
category.
3) A Pareto Diagram shows counts for each category as a fraction of total counts.
4) Pareto Diagrams, when compared with the one before corrective actions and the other after, show the
changes and effects brought by the measures.
Note) When making a Pareto-Diagram-to-Pareto-Diagram comparison, keep the scale of the vertical
column y axis the same among the diagrams to be compared with.
10 Constructing a Pareto Diagram

[Step 1] Decide what information you want to get from the diagram.

Quality Defective item, failure item, complain item, etc.

Cost Cost item, kind of material, etc.
Delivery Reason for delay, number of days delayed, etc.
Morale Absent days, number of suggestions for improvement made, etc.
Manning Person, years of experience, gender, etc.
Equipment, tools Machine, fixture, measuring instrument,work line, etc.
Material Material, brand, lot, maker, etc.
Operation Operating condition, time, size, etc.
Environment Weather, region, season, etc.

[Step 2] Decide how to collect data.

Decide what period of time the diagram will cover.

Key Points
①Speed is required in problem-solving. Make the best use of the data already exist, where
appropriate.
②Putting the way occurrences are taking place into perspective, systematically set a period
of time that the diagram covers, for example, a week, a month, etc., so that plotted bars
may not be extremely biased.
③When collecting numerical data, check details and causes of occurrences.
It pays when taking corrective/preventive measures.
④Use “quantitative” data.
Nonconformance occurrences, nonconforming parts, frequency of stoppage, frequency of
detection, frequency of failure, cost, loss amount, sales amount, etc.

[Step 3] Sort and total the data in accordance with each category (content and cause).

<<Reference>>
The Pareto principle is derived from the work of Vilfredo Pareto, a 19 th century Italian
economist who studied the distributions of wealth. He concluded that a fairly consistent
minority of people controlled the large majority of a social wealth.
Later, Dr. Joseph M. Juran, an American QC authority, applied the Pareto principle to quality
control and popularized it as a powerful QC tool.
 Exercise 11

[Step 1] Decide what information you want to get from the diagram.

Decide what categories you will use to group items.

Check number of defective units for each category item and determine where to focus efforts.

[Step 2] Decide how to collect data.

Decide what period of time the diagram will cover.

Data: Use the inspection data
Period: 3 months

[Step 3] Sort and total the data in accordance with each category (content and cause).

Put the data (the inspection results) together in a spreadsheet.

Finished Parts Inspection spread sheet (Primary Defects)

Part No. BPR6ES Period Mar 04 – May 31

Inspector: Takeshita

Month
Defect item
3 4 5 Total
Damage 126 48 62 236
Burrs 26 20 22 68
Plating 30 27 32 89
Contamination 28 68 111 207
Chatter marks 1 0 0 1
Chipping 0 0 1 1
Total 212 163 227 602

Remarks

Nothing to mention in particular

12 Constructing a Pareto Diagram

[Step 4] Sort data.

Calculate cumulative counts.

Key Points

①Put the categories in order, placing the one with the largest count first, then the next largest, etc.
Enter the count for each category.
②Put the categories with insignificant counts together into one category labeled “other” and place
it last.
③Calculate cumulative counts starting with the category with the largest count and continuing
with each category that follows in descending order.

[Step 5] Calculate cumulative share.

Cumulative counts
Cumulative share =  100％
Total

[Step 6] Label vertical and horizontal axes.

Key Points
①Label the left-hand vertical axis with frequency of occurrence. Label scales, spacing them in
equal intervals from zero to a number equal to or just larger than the total of all counts.
②Label the horizontal axis with problem categories in descending order from left to right.
③Arrange the scales so that “the length of vertical axis vs. the length of horizontal axis” will be
1:1 to 2:1 (a typical Pareto Diagram is near-square in shape) for good visibility.

[Step 7] Create a bar chart for each category

Key Points
①Plot bars so that the height of bar for each category is equal to the count for that category.
②Draw each bar with the same width, without spacing between bars.
 Exercise 13

[Step 4] Calculate cumulative counts.

Defectiv Cumulative Cumulative
No. Category e units counts share
No.1 236
1 Damage 236 236
No.2 236＋207＝443
2 Contamination 207 443
No.3 443＋ 89＝532
3 Plating 89 532
No.4 532＋ 68＝600
4 Burrs 68 600
No.5 600＋ 2＝602
5 Others 2 602
Total 602 –

[Step 5] Calculate cumulative share. Defectiv Cumulative Cumulative

No. Category e units counts share
236 1 Damage 236 236 39.2%
No.1  100  39.2%
602 2 Contamination 207 443 73.6%
443
No.2  100  73.6% 3 Plating 89 532 88.4%
602
532 4 Burrs 68 600 99.7%
No.3  100  88.4%
602
5 Others 2 602 100.0%
Total 602 – –

[Steps 6 & 7] Plot a bar chart for each category.

(Pieces)
600

500

400
Defective unit

300

200

100

0
Damage

Contamination

Plating

Burrs

Others
14 Constructing a Pareto Diagram

[Step 8] Add a right-hand vertical axis. Label it with percentage scales.

(The scales should range from 0 to 100%, lining up the 100% with the grand total on the right-hand
vertical axis).
①The starting point of line graph should be labeled with 0%, the ending point with 100%.
②Divide “0~100%” by 2~5 for equal intervals. Mark scales and label with the corresponding
percentage.
<<Example>>
Divide “0~100%” by 4 and label 25, 50, ….100%.

[Step 9] Add a cumulative line.

For each category, put a dot as high as the cumulative total on a vertical phantom line drawn in line
with the right edge of the bar of that category. Connect the dots with lines to create a line graph.

[Step 10] Enter necessary items.

Title, period, total of data counts, process name, author, etc.

[Step 11] Analyze

Key Points
- Evaluate each category as a percentage of the total counts in order to decide the few
significant categories against which you’ll take corrective actions.
- It is effective to take actions against the significant few items that will amount to 60 to 70%
of the total count.
 Exercise 15

[Step 8] Add a right-hand vertical axis. Label it with percentage scales

(Pieces) 600 100(%)

500
75
400
Defective unit
300 50

200
25
100

0 0
Damage

Contamination

Plating

Burrs

[Steps 9, 10] Add a cumulative line. Enter necessary items. Others

Defective Finished Plug: Pareto Diagram (Primary Defects)

Part No. BPR6ES Period Mar.04 – May 31 Inspector: Takeshita
n = 602
100(%)
(Pieces) 600

500
75
400
Defective unit

300 50

200
25
100

0 0
Damage

Contamination

Plating

Burrs

Others

[Step 11] Analyze

The two problem categories, “damage” and “contamination,” account for 74% of
the total defective units. It is effective to concentrate on “damage” and
16

“contamination” for corrective actions.