Knowledge sharing on

7 QC TOOLS
By
Chandrashekhar Chaudhary
Pravin Bhalerao
QC TOOLS

Why do we need QC Tools ?

Improve Business Results
Improve QCDSM
Take Appropriate Decisions.
Decisions Based On Fact
Decisions Based On Opinion

What is the difference between Fact and Opinion?

QC TOOLS…

Difference between Fact & Opinion is

Fact is Based on Data
Opinion is based on Judgment / Experience.
Please read the statement as stated below & find out what is fact & what is
opinion ?
During this Summer Season, The temperature has gone up which led to
increase in power consumption drastically from 9000MW to 20000MW.
Due to this the Sale of Air Conditioner has increased two fold from 50,000
to 1,00,000 nos. of Kelvinator Industries in comparison to last year.

Fact : 9,000MW to 20,000MW, Sale of Air Conditioner increased two fold

from 50, 000 to 1,00,000Nos.
Opinion : The Temperature has gone up, Increase in power consumption
drastically

Conclusion : QC Tools are used to take appropriate decisions based on

DATA
QC TOOLS…
Why do we need QC Tools ?

 Systematic method to collect the data

 Stratifying the data with various ways

 Identify the vital few from trivial many (Prioritization to have

focused effort )

 Identifying the various causes of the problem, understanding

clear linkage between primary, secondary causes etc.

 Understanding the behavior of the process

 Understanding the relation between parameters (output

characteristics with process & input)

 Identifying the real root cause of the problem

QC TOOLS…
Why do we need QC Tools ?..........
 Good Parts & defective Parts are produced from the same process. So,
why some parts are defective & some are O.K.
 Defects are due to Variation.
 Variation in what?
 Variation in Input (Incoming Material)
 Variation in Process ( Man, Machine, Method, Tool, Environment,
Measurement System etc.)

 QC tools are simple & effective to solve problems. It is data

based approach & supports to take appropriate decision.
Through QC Tools, we understand the behavior of the process.
Relation of Output with Input and process. We can analyze the
data & come up with meaningful conclusion.
Types of QC Tools
Basic 7 QC Tools – Problem Solving Tool
New 7QC Tools – Management Decision Making Tool
QC TOOLS…

Best Methods adopted for problem solving …..?

Simple Very Difficult

Implementation is not Breakthrough
easy improvement
High

Best Suited for QC Circle Companywide

Best Suited for CFT / QC
Story / 6Sigma
Effort

Very Simple Difficult

Implementation is easy Breakthrough
Best Suited for Improvement
Less

Suggestions Companywide
Best Suited for
Supervisory Improvement
team

Less High
Complexity
BASIC 7 QC TOOLS
7 QC TOOLS

What are the 7 QC Tools?

1) Check Sheet (Systematic Method to collect data)
2) Flow Charts (Pictorial representations of a process)
3) Pareto Analysis (Identifying the Vital few from Useful many)
4) Cause & Effect Diagram (Identifying the various causes of the problem ,
understanding the clear linkage between primary, secondary , tertiary
causes etc.)
5) Graph & Control Charts (Understanding the behavior of the process
Pictorially)
6) Histogram (Understanding the Distribution of the Output / characteristic
pictorially)
7) Scatter Diagram (Understanding the relation between two or more
parameters)

*** Stratification (Stratifying the data in various ways)

1. CHECKSHEET

 Purpose
 Simplification of Data Collection process
 Reduce recording errors
 Ease of analysis

 Types of Check Sheet

 Production process distribution Check sheet (Variable / Attribute data) –
Collects data in easy manner & distribution can be seen.
 Defective Item Check sheet – Defect wise data collected to prioritize.
 Defect Location Check sheet – Helps in understanding the location where
more defects, such as Dents / scratches / Crack / Necking / wrinkles etc., are
generated.
 Defect Cause check sheet – Shift wise , Machine wise , Operator wise etc.
data is gathered. It provides clue for solving the problem
1. CHECKSHEET…

Check Sheet for Production Process Distribution

Spec : 30.5 ± 0.006 mm
CHECKS
Spec. Deviation Frequency
1 2 3 4 5 6 7 8 9 10
-10
-9
-8 1
-7 2
-6 2
-5 3
-4 3
-3 4
-2 5
-1 6
30.5 0 7
1 5
2 4
3 3
4 3
5 2
6 2
7 1
8
9
10
1. CHECKSHEET…

Defective Item Check Sheet – Incoming Inspection

Check Sheet
Product _____________ Date ___________________________________
Supplier ______________ Inspector Name_________________________
Sheet Grade __________ Lot Size ______________________________
S# Defect Type Frequency Marks Frequency
1 Cracks I 1
2 Wrinkles IIII 5
3 Flange Short IIII IIII 10
4 Thinning IIII IIII IIII 15
5 Burrs IIII IIII IIII IIII II 22
6 Rust IIII IIII IIII IIII 20
7 Pitting IIII IIII 10
8 Folding IIII I 6
9 Dents IIII 5
Total Defects 94
IIII IIII IIII IIII
Total Defectives
IIII IIII IIII IIII II
42
1. CHECKSHEET…

Defect Location Check Sheet – Draw Operation during stamping

CRCA Sheet
1 2 3 4

5 6 7 8

9 10 11 12

13 14 15 16

Location
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Frequency
Defect
Crack IIII II IIII II III III III IIII 28

Wrinkle III II IIII IIII IIII IIII III III II 32

Dents IIII III II III II II II II III II 26

Frequency

11 0 2 2 7 5 3 7 11 2 3 10 3 9 2 7
 1. CHECKSHEET…

Defect Cause check Sheet

Mon Tue Wed Thu Fri Sat

Equipment Worker
AM PM AM PM AM PM AM PM AM PM AM PM

Machine - 1

Machine - 2

Legend Crack Wrinkles Burr Blank Short

*** STRATIFICATION

STRATIFICATION
Stratification means Classification / Grouping of Data based on
Particular Features
e.g.

Skill, experience, age. Union,

Operator wise
Qualification

Manufacturer, Type, Date of

Material Wise
arrival, Date of Use, Batch

Day, Night, Shift, Working Days,

Time wise
Holidays, Week

Type of Machine, Machine No. ,

Machine wise
Machine Capacity
 *** STRATIFICATION…

Officers in M&M – Stratification – Department wise / Age wise / Sex wise etc.

Mahindra Group
Age 18~25 26~40 41~60 Total
Deptt. Male Female Total Male Female Total Male Female Total Male Female Sub Total
CQA 2 0 2 3 0 3 1 0 1 6 0 6
SQA 25 1 26 3 0 3 1 0 1 29 1 30
PVT 22 2 24 4 0 4 2 0 2 28 2 30
MM 26 2 28 10 0 10 2 0 2 38 2 40
BTCF 58 0 58 20 0 20 2 0 2 80 0 80
STCF 98 2 100 17 0 17 3 0 3 118 2 120
UV Body 34 0 34 8 0 8 2 0 2 44 0 44
Scorpio Body 28 0 28 8 0 8 2 0 2 38 0 38
Xylo Body 26 0 26 3 0 3 1 0 1 30 0 30
Paint 32 0 32 8 0 8 2 0 2 42 0 42
R&D 24 3 27 8 0 8 5 0 5 37 3 40
2. FLOW CHARTS

Flow Charts
 Flowcharts are pictorial representations of a process using simple
Symbols & Arrows. By breaking the process down into its constituent
steps, flowcharts can be useful in identifying where errors are likely to be
found in the system
 The first structured method for documenting process flow, the "flow
process chart", was introduced by Frank Gilbreth to members of the
American Society of Mechanical Engineers (ASME) in 1921 .
 A flowchart shows the activities in the process & the relationship between
them.
 Flowchart lists every step involved in the manufacture of a product or
delivery of a service.
 Excellent tool to identify wasteful (NVA) actions.
 It documents the process completely.
 Systematic approach to reveal possible sources of Quality & Productivity
Problems.
 Commonly used to promote improvements.
2. FLOW CHARTS

Use of Flow Charts:

 Define and analyze processes;
 Build a step-by-step picture of the process for analysis,
discussion, or communication; and
 Define, standardize or find areas for improvement in a process.
 Bifurcate Value Added & Non-Value Added activities.

SYMBOLS :
 Elongated circles, which signify the start or end of a process
 Rectangles, which show instructions or actions; and
 Diamonds, which show decisions that must be made

Start Activity Decision End

2. FLOW CHARTS

Why use it?

The flow chart helps you analyze the number of steps and the time
required for each step, to detect bottlenecks, unnecessary steps,
repetitions, and other obstacles

When to use it?

When you want to describe activities, identify problem, identify
causes of problems, detect ―bottlenecks‖, and define indicators
2. FLOW CHARTS

Use of Flow Charts:

 Define and analyze processes;
 Build a step-by-step picture of the process for analysis,
discussion, or communication; and
 Define, standardize or find areas for improvement in a process.
 Bifurcate Value Added & Non-Value Added activities.

Key Points:
 Flow charts are simple diagrams that map out a process so that
it can easily be communicated to other people.
 To draw a flowchart, brainstorm the tasks and decisions made
during a process, and write them down in order.
 Then map these out in flow chart format using appropriate
symbols for the start and end of a process, for actions to be taken
and for decisions to be made.
 Finally, challenge your flow chart to make sure that it's an
accurate representation of the process, and that it represents the
most efficient way of doing the job.
2. FLOW CHARTS
Example-1: The example below shows part of a simple flow chart which helps
receptionists route incoming phone calls to the correct department in a company:
2. FLOW CHARTS
Example-2: A Simple Flow Chart for Dealing with a Non Functioning Lamp

Lamp Doesn't
Work

No
Lamp Plugged Plug In Lamp
In?

Yes

Yes
Bulb Burnt Replace Bulb
Out?

No

Buy New
Lamp
2. FLOW CHARTS
Example-3: A Simple Flow Chart for diagnosis of customer complaints at
dealership
2. FLOW CHARTS

Some of the Widely used Flow Chart Symbols are as follows :

Operation
Inspection

Transportation

Delay

Storage

Operation & Inspection

Note : Other than above symbols, There are many Symbols also exists. You
may look into Website for details.
2. FLOW CHARTS

Benefits of Flow Charts are as follows :

 Create Visual map of process.
 To identify time lags & NVA steps.
 Identify process that need improvement.
 Determine major & minor inputs in the process.
 Promotes process understanding.
 Depicts customer - supplier relationship
3. PARETO ANALYSIS

PARETO ANALYSIS
 Developed By Wilfredo Pareto an Italian economist and applied by Dr. JM
Juran.
 Used for Prioritization by 80~20 Rule.
 Differentiates ―Vital Few‖ from ―Useful Many‖.

 Pareto Principle states that

 20% of the Defects contributes to 80% of the Rejection, OR,
 20% of the Causes contributes to 80% of the Problems, OR
 A Few Causes Lead to Many Defects & Many Causes leads to Few Defects.
 The Few Causes that Lead to many defects are The ―VITAL FEW‖.
 The many causes that lead to Few defects are the ―USEFUL MANY‖
 Get to the Biggest Problems first – Solve Vital Few.
3. PARETO ANALYSIS

When to Use PARETO ANALYSIS

 Starter to Problem solving - What to solve?

 To break big problems into smaller problems.

 To prioritize high impact issues - Where to focus?

 Systematic analysis of causes based on magnitude.

 Allows better use of limited resources.

3. PARETO ANALYSIS…
 Key Steps for Plotting Pareto Diagram
 Arrange the data in Descending Order.
 Mention Others at the end (Combining all other minor defects)
 Calculate the Total no. of defect & Calculate Cumulative %.
 Select the Scale in Such a way that Total value of defect in Y axis of Left
side is exactly matching with 100% on the right side of Y axis (Excel sheet
– Draw Chart – Custom – Line – Column on 2Axis).
 Plot the Pareto Chart as shown in example – Bar Chart for Numbers & Line
chart for Cumulative %.
 Draw horizontal line for 80% and identify the point of intersection on
cumulative graph.
 Draw the vertical line from the point of intersection.
 Identify Left side of vertical line as ―VITAL FEW‖ and right side as ―USEFUL
MANY‖. Action to be initiated on VITAL FEW.
 Pareto of vital few may be analyzed further to next level.
 If a problem / defect is expected to have simple solution, it should be
tackled right away even though it is relatively of lesser importance.
3. PARETO ANALYSIS…
3. PARETO ANALYSIS…
3. PARETO ANALYSIS…
Actual data
Defect Qty.
Pareto Analysis - Scorpio Body Shop - Jun'09
Nut Miss 50
Part Miss 5
Crack 12 284 100
Wrinkle 26 90
Burr 18
80
Blow Holes 28 213

Rej. Qty. (In Nos.)

70
Spot Dent 30
Spot Failure 55 60

Cum %
Undercut 15 142 Vital Few Useful many 50
Spot Puncture 20 40
Flange Short 5 30
Wrong part 8 71
20
Wrong Position 10
Thread crack 2 10
0 0

Burr

Undercut

Crack
Spot Puncture
Spot Dent

Wrong part
Nut Miss
Spot Failure

Thread crack
Part Miss

Flange Short
Wrinkle
Blow Holes

Wrong Position
Defect Qty. Cum %
Spot Failure 55 19.37
Nut Miss 50 36.97
Spot Dent 30 47.54
Type of defects
Blow Holes 28 57.39
Wrinkle 26 66.55
Spot Puncture 20 73.59
Burr 18 79.93
Undercut 15 85.21
Crack 12 89.44
Wrong Position 10 92.96
Wrong part 8 95.77
Part Miss 5 97.54
Flange Short 5 99.30
Thread crack 2 100.00
Total 284
3. PARETO ANALYSIS…

How to Plot Pareto Graph in New Version of Windows :

Nature Of Defect Qty. % Cum Sum % Contribution

Spatter 50 26.60 26.60
Crack 40 47.87 21.28
Spot Puncture 30 63.83 15.96
Pin Hole 25 77.13 13.30
Thinning 25 90.43 13.30
Blow Holes 10 95.74 5.32
Dents 5 98.40 2.66
Damages 3 100.00 1.60
Total 188
3. PARETO ANALYSIS…
Step-1: Select QTY. & % Cum Sum Columns & Plot Line Graph
Route : Insert - Line - Line with Markers
3. PARETO ANALYSIS…
Step-2: Select Qty. Line on Graph & convert Line into Column
Route : Pick Qty. Line Graph - Change Series Chart Type - Column - Clustered
column
3. PARETO ANALYSIS…
Step-3: Select the Legend entry & delete it
Route : Pick Series block – right click - delete
3. PARETO ANALYSIS…
Step-4: Select the % Cum Sum Line Graph & Add Secondary Axis
Route : Pick % Cum Sum Line Graph - Format Data Series - Series Options -
Secondary Axis - Close
3. PARETO ANALYSIS…
Step-5: Select the Primary Axis & ensure that Maximum limit
Equals to Total Qty.
Route : Pick Left Side Primary Axis - Format Axis - Axis Options -
Minimum/Maximum Values - Fixed - Put Value - Close
3. PARETO ANALYSIS…
Step-6: Select the Secondary Vertical Axis & ensure that Maximum limit
Equals to 100% ( Total Qty Value corresponds to 100%).
Route : Pick Right Side Vertical / Secondary Axis - Format Axis - Axis
Options - Minimum/Maximum Values - Fixed - Put Value - Close
3. PARETO ANALYSIS…
Step-7: Remove Decimal Digits from % Cum Sum Axis Values
Route : Pick Secondary Vertical Axis - Format Axis - Number - Decimal Places
- Select "0" - OK
3. PARETO ANALYSIS…
Step-8: Add X- Axis data level
Route : Pick Chart - Select Data Source - Horizontal Category - Axis Labels -
Edit - Pick Column "Nature Of Defects" - OK
3. PARETO ANALYSIS…
Step-9: Reduce the size of Bullet on % Cum. Sum. Graph
Route : Pick % Cum. Sum Line Graph – Format Data Series – Marker Options.
3. PARETO ANALYSIS…
Step-10: Select Chart & Put Chart Titles/Axis Titles
Route : Pick Chart - Layout - Chart Title/Axis Titles
4. CAUSE & EFFECT DIAGRAM

Take away the cause, and the effect ceases.

-Cervantes
4. CAUSE & EFFECT DIAGRAM
CAUSE & EFFECT DIAGRAM
 Developed by Professor ISHIKAWA, hence also known as Ishikawa Diagram.
 Also known as Fish Bone Diagram due to its shape & structure.
 This is a diagram which shows a relationship between a Quality
Characteristic and Factors affecting it.
"As much as 95% of Quality related problems in the factory can be solved
with Seven fundamental quantitative tools.‖ - Kaoru Ishikawa

Brief of Prof. ISHIKAWA :

 Born on 13th July‘1915 in Tokyo Japan
 1939: Engineering degree in applied chemistry
from University of Tokyo
 Conceived Cause & Effect diagram for root cause
analysis (Fish bone diagram / Ishikawa diagram)
 1949:Joined the Japanese Union of Scientists
and Engineers (JUSE)
 Introduced the concept of Quality circle Kaoru Ishikawa
(1915-1989)
 Enabler in transforming Japanese industrial sector
from cheap producers to world class standards
4. CAUSE & EFFECT DIAGRAM
Key Steps for Preparing Cause & Effect Diagram
 Identify the concerned team members for Effective Brainstorming
 Brainstorming is a tool used by teams for creative exploration of
options in an environment of free criticism. It provides creative and
unrestricted exploration of options, ideas, causes or solutions.
• Write down for what you are doing Brainstorming.
• Ask People to think individually.
• Collect as many ideas as possible.
• Absolutely no discussion on ideas.
• Do not criticize / judge any idea.
• Set the time limit.
• Each one calls out idea and facilitator to record them on board.
 After Brainstorming & listing the Causes, write the effect as Head of the
fish and Causes as Bones.
 Connect Main Causes to effect & Sub Causes to Main Cause carefully.
4. CAUSE & EFFECT DIAGRAM

 Benefits of Brainstorming

 Creativity

 Large number of ideas

 Involvement of team members

 Sense of ownership in decisions

 Input to other tools

4. CAUSE & EFFECT DIAGRAM

 Used to explore potential & real causes.

 Compares relative importance of each cause

 Helps to identify root cause.

 Increases knowledge of a process, its factors etc.,

 Identifies areas of further data collection.

4. CAUSE & EFFECT DIAGRAM

Method for Construction of Cause & Effect Diagram

• State the problem, place it in a box on the right side
• Draw a horizontal arrow to the problem box
• Write the primary causes which affect the problem as Big Bone.
You can use traditional main categories of factors* or suspected
categories above and below the line, and connect them to the
main line
• Write Secondary causes which affect the primary causes as
medium size bones. And tertiary causes which affect medium
sized bones.
• Assign importance to each factor and mark those which have
significant impact on the problem
4. CAUSE & EFFECT DIAGRAM

Man Machine

Effect

Materials Method
(Environment)
4. CAUSE & EFFECT DIAGRAM

BENEFITS :

Helps determine root causes

 Encourages group participation

 Uses an orderly, easy-to-read format

 Indicates possible causes of variation

 Increases process knowledge

 Identifies areas for collecting data

4. CAUSE & EFFECT DIAGRAM
Limitations :
 There is no logical consistency in the relationship between effect,
causes and sub-causes.

 Enough probing is not done to identify all probable root causes.

 Often multiple Whys are not asked and root causes are not fully
identified,

 The causes shown by the cause and effect diagram are only probable
causes till they are validated with data.

 Often people assume them to be real root causes and directly start
taking action against them without validating them with data.

 This will only lead to wasted efforts and cause frustration and may
not solve the problem.
4. CAUSE & EFFECT DIAGRAM
 4. CAUSE & EFFECT DIAGRAM…
Example-1:
Carbon Test not Man Material Variation in Sheet Thickness
done

Parameter Setting Wrong BOM

Not defined in Disturbed Oily Sheet
SOP
Wrong Supply
No Poka- Yoke Wrong
Improper Gun Sheet
Alignment Panel Surface
Not defined in SOP not Clean
High Carbon %
Tip Not Dressed
Poor Preservation
Not defined in SOP
Chisel & Hammer Paper adhered High Hardness
testing not done
Rusty Sheet
New Operator
Spot
Clamps Not No in Built
Failure
working No UTM Test Stabilizer
Wrong Capacity Not
position of Inadequate
Destructive defined in SOP
clamps Fluctuation in
testing Voltage
Improper
clamping Chisel test frequency M/C Gun approach
Wrong mode inadequate running of less
of holding Genset
Inadequate the Gun Inadequate tip
clamps Force Wrong
Welding Gun
Untrained Capacity of
Poor Maint. Operator Compressor Inadequate Poor Cooling of
less Line Air electrode
Poor resting of Poor Gun Pressure
part on Pads Approach
No Water Flow
Machine
Fixtures Method
4. CAUSE & EFFECT DIAGRAM…
Example-2:
4. CAUSE & EFFECT DIAGRAM…
Example-3:
4. CAUSE & EFFECT DIAGRAM…
Example-4:

Material Machine Man

Grease quality poor Locktite not Greasing not done

Tube flange Applied on
broken threads Oil seal not fitted
Shaft dimension Improper
Brg. Dimension Wrong endplay torquing
not ok Not ok
Crimping
Foreign particle not done
In grease
Axle Shaft
Single piece diff case Overloading Breakage
& 2 pinion mate done

Loosening of nut Shaft threads

under load strength less
Grease on shaft Greasing not done
Wrong threads during servicing
Overloading not interference
considered

Design Method Usage

4. CAUSE & EFFECT DIAGRAM…
Example-5:
5. GRAPH & CONTROL CHARTS

Brief History:
A control chart was first proposed in 1924 by W.A Shewart,
who belonged to the Bell Telephone Laboratories, with a
view to eliminating an abnormal variation by distinguishing
variations due to assignable causes from those due to chance
causes.

Definition of Graphs:
Graphs are pictorial depiction of a set of Data for Ease of
Understanding, Interpretation & Analysis. E.g. Line Chart ,
Bar Graph, Pie Chart, Radar Chart etc.
5. GRAPH & CONTROL CHARTS
Example of Graphs
Bar Chart - Defect W ise
Nature of Defects Incidences
60 50
No. of Incidences

50 Scratch 50
40
40 30
25 Dents 25
30 20
20 10 Nut Miss 40
10
0 Crack 30
Scratch Dents Nu t Miss Crack Spot Blow
Spot Failure 20
Failu re Holes
Natu re of Defects Blow Holes 10

Defect Trend for Scratch

Line Chart - Monthwise distribution of Defect
Month & Year Incidences
" Scratch"
Jan'10 50
100 79
No. of Incidences

80 66 Feb'10 48
50 48
60 38 Mar'10 32
32
40
20 Apr'10 66
0
Jan'10 Feb'10 Mar'10 Apr'10 May'10 Jun'10
May'10 79
Month & Year
Jun'10 38
5. GRAPH & CONTROL CHARTS

Definition of Control Charts:

 Statistical Tool for Monitoring & Improving Quality.
 Distinguishes Between Common & Special Causes.
 Measures Consistency of a Machine or a Process.

UCL

LCL

When to use Control Chart:

 Visual display for Process output.
 To monitor, control & improve process performance.
 To identify variation at its source.
5. GRAPH & CONTROL CHARTS

WHY GRAPH & CONTROL CHARTS (SPC)

 To understand the behavior of the process through Graphs &
Control Charts & accordingly initiating actions to improve the
process.
 Philosophy of SPC
 No two things are identical
 Variation is inevitable
 Variation is measurable
 Variation is Predictable
 Reduction in Variation will reduce the defects and improve the
quality.
 Listen Voice of the process as Process cannot understand
Specifications.
5. GRAPH & CONTROL CHARTS…
 Goal Post mentality : If the Parameter is within Specifications, we say that
the characteristic is O.K. and loss will occur only when it will go beyond
Specifications.
 But in Actual, as we go away from the mean, the loss (In Terms of Money) is
going on increasing. This can be easily understood by Taguchi Loss
Function.

LSL µ USL

Normal
Distribution
Curve

Taguchi Loss
Curve
3 3
5. GRAPH & CONTROL CHARTS…
 Variations are due to :
 Chance Causes / Common Causes
 Assignable Causes / Special Causes

Parameter Common Cause Special Cause

Nature System Deficiency Process failure

Scope Of All production parts are Only a few production parts affected
influence affected equally similarly
Sources Many Small Sources A few Major Sources

Process Stable & Predictable Sporadic & unpredictable

Behaviour
% of Total 85% ~ 95% 5% ~ 15%
Variation
Responsibility Management Supervisor / Operator

e.g. In Built Process Design e.g. Production continued even though

Jigs & Fixtures, Machine Tool get broken, M/C running with
Selection, Parameter setting, worn out bearings, Tool in use even
Tools & Dies, Process Control after its life is over, Change in 4M,
mechanisms Measurement Error
5. GRAPH & CONTROL CHARTS…

Special Causes of Variation :If special causes of variation are

present, the output of a process forms a distribution that is not
stable over time. It is not predictable.
Variation from special cause is unpredictable.
Variation from special cause is unstable
The mean changes over time
The spread changes over time
The shape varies over time

Common Causes of Variation : If only common causes of variation

are present, the output of a process forms a distribution that is
stable over time. It is predictable.
 Variation from common cause is Predictable.
 Variation from common cause is Stable
The mean is unchanged over time
The spread is unchanged over time
The shape is unchanged over time
5. GRAPH & CONTROL CHARTS…
5. GRAPH & CONTROL CHARTS…
Statistical Process Control (SPC) - Process is said to be Stable if the Process is free from
Special Cause. It is also called Process Under Statistical Control.
40.00 40.00
35.00 35.00
30.00 30.00
25.00 25.00
20.00
20.00
15.00
15.00
10.00
10.00
5.00
5.00
0.00
1 2 3 4 5 6 7 0.00
1 2 3 4 5 6 7

Increasing Trend – Process Unstable Decreasing Trend – Process Unstable

20.00 19.00
18.50
18.00
18.00
16.00 17.50

14.00 17.00
16.50
12.00
16.00

10.00 15.50
15.00
8.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14

7 Continuous Points on one side of Process Stable

Mean - Process Unstable
6. HISTOGRAM

HISTOGRAM OVERVIEW
 Histogram is a tool to Pictorially demonstrate the distribution of the
Output / characteristic
 The characteristics of the frequency distribution are shown more clearly
when results are plotted in form of Block Diagram.
 The horizontal axis is divided into segment corresponding to the ranges of
the group.
 On each segment a rectangle is constructed whose height is proportional to
the frequency in the group.
 Higher bar represents more data values.
 Lower bar represents less data values.
 It Provides following information about the Process :
 Shape of the Distribution
 Approximate Average / Mean
 Spread
 Level of Conformance to Specification.
6. HISTOGRAM

HISTOGRAM DEFINATION
 A vertical bar chart that depicts distribution of data.
 It is a one time snap shot of a process performance.
 A tool to determine the frequency of occurrence of data.
 A graphical information communicating tool.
 Compares process results with specification limits.

WHEN TO USE HISTOGRAM

 To summarize large data sets graphically
 To compare measurements to specifications
 To communicate information to the team
 Assist in decision making
6. HISTOGRAM…

Key Steps for drawing Histogram :

 Step-1 : Obtain a Set of 50~100 observations
6. HISTOGRAM…
Key Steps for drawing Histogram :
 Step-2 : Obtain Maximum Value & Minimum Value. Find Out the Range.
R = Maximum Value – Minimum Value = 2.545-2.502 = 0.043.
6. HISTOGRAM…
Key Steps for drawing Histogram :
Step-3 : Determine the No. of Bins (k)

 The No. of Bins = SQRT (Total No. of observations). It is rounded

off to nearest integer.
 k = SQRT(90) = 9.48  9 (Rounded Off).

Step-4 : Determine the Class Interval (c)

 Class interval is determined so that the range will include the

maximum & minimum of the observed values.
 Class Interval = Range / No. of Bins.
 c = 0.043/9 = 0.005

Note : No. of Bins should contain 5~20 Class Intervals of equal

width.
6. HISTOGRAM…
Key Steps for drawing Histogram :
Step-5 : Determine the Class Boundaries

 Determine the starting point of the Class Interval.

 Starting Point = Lowest Observed Value – (Least Count/2) OR,
 Starting Point = Lowest Value.
 In this Case, Let the Smallest Value = 2.5005
 To Obtain the Class Boundary of the First Bin, add class interval to
Starting Point.
 Determine the boundaries of Each Bin so that they include the
smallest & the Largest Values & write these down on the frequency
table. Note that the Last Bin must contain the Largest observed value.

Step-6: Calculate the Mid Point of Each Class

 Mid Point = Sum of upper & Lower Boundaries /2.

 In this Case,
• Mid Point of 1st Class = (2.5005+2.5055)/2 = 2.503
• Mid Point of 2nd Class = (2.5055+2.5105)/2 = 2.508
6. HISTOGRAM…
e.g.
Thickness of Lens
Sample No.
Results of Measuremen
1~10 2.51 2.517 2.522 2.522 2.51 2.511 2.519 2.532 2.543 2.525
11~20 2.527 2.536 2.506 2.541 2.512 2.515 2.521 2.536 2.529 2.524
21~30 2.529 2.523 2.523 2.523 2.519 2.528 2.543 2.538 2.518 2.534
31~40 2.52 2.514 2.512 2.534 2.526 2.53 2.532 2.526 2.523 2.52
41~50 2.535 2.523 2.526 2.525 2.532 2.522 2.502 2.53 2.522 2.514
51~60 2.533 2.51 2.542 2.524 2.53 2.521 2.522 2.535 2.54 2.528
61~70 2.525 2.515 2.52 2.519 2.526 2.527 2.522 2.542 2.54 2.528
71~80 2.531 2.545 2.524 2.522 2.52 2.519 2.519 2.529 2.522 2.513
81~90 2.518 2.527 2.511 2.519 2.531 2.527 2.529 2.528 2.519 2.521

Minimum Value = 2.502

Maximum Value = 2.545
Range ( R )= 0.043
No. of bins = n = 90 = 9.48683  9
R
Class interval = = 0.00478  0.005
n

Starting Point = (Minimum Value - Least Count/2), OR, Less Than the Smallest value = 2.5005
1st Class Boundary = 2.5005+Class interval = 2.5005+0.005 = 2.5050 i.e. 2.5005~2.5055
2nd Class boundary = 2.5055+0.005 = 2.5105 i.e. 2.5055~2.5105
3rd Class boundary = 2.5105+0.005 = 2.5155 i.e. 2.5105~2.5155
4th Class boundary = 2.5155+0.005 = 2.5205 i.e. 2.5155~2.5205
5th Class boundary = 2.5205+0.005 = 2.5255 i.e. 2.5205~2.5255
6th Class boundary = 2.5255+0.005 = 2.5305 i.e. 2.5255~2.5305
7th Class boundary = 2.5305+0.005 = 2.5355 i.e. 2.5305~2.5355
8th Class boundary = 2.5355+0.005 = 2.5405 i.e. 2.5355~2.5405
9th Class boundary = 2.5405+0.005 = 2.5455 i.e. 2.5405~2.5455

Calculate Mid Point for each class = (Upper Boundary + Lower boundary)/2
6. HISTOGRAM…
Key Steps for drawing Histogram :

Step-7 : Prepare frequency Table

 Count the no. of values falling in each class interval & put Tally
Mark & Then count the Frequency

Step-8 : Mark the Horizontal Axis with Class Boundary Values.

Step-9 : Mark the Vertical Axis with a Frequency Scale.

Step-10 : Erect the rectangles over the class interval having area
proportion to the frequencies.

Step-11 : Draw a line / Smooth Curve on the Histogram to Represent

the Mean & Standard Deviation. Expand the Bars / rectangles so
that they touch each other.
6. HISTOGRAM…

25

Frequency 20

15

10

0
2.503 2.508 2.513 2.518 2.523 2.528 2.533 2.538 2.543
Class Interval

S# Class Mid Point Frequency Marks Frequency

1 2.5005 ~ 2.5055 2.503 I 1
2 2.5055~2.5105 2.508 IIII 5
3 2.5105~2.5155 2.513 IIII IIII 10
4 2.5155~2.5205 2.518 IIII IIII IIII 15
5 2.5205~2.5255 2.523 IIII IIII IIII IIII II 22
6 2.5255~2.5305 2.528 IIII IIII IIII IIII 20
7 2.5305~2.5355 2.533 IIII IIII 10
8 2.5355~2.5405 2.538 IIII I 6
9 2.5405~2.5455 2.543 IIII 5
6. HISTOGRAM…
6. HISTOGRAM…
• Applicability of Histogram :
 To display large amount of data in simple chart form and analyze the
process.
 To tell relative frequency of Occurrence.
 To easily see the distribution pattern of the data.
 To see there is variation in data.
 To make future predictions based on data.
 To compare the improvement (before & After).
 To compare the two processes.

• Interpretation of Histogram :
 Look at the specifications – it indicates Acceptance / Rejection or Non-
conformance.
 Look at the Peak – It represents Process Setting.
 Look at the width – it represents Variability in Process.
 Look at the Shape – it represents Predictability.
6. HISTOGRAM…

Types of Distribution
 The shape of the distribution gives a more elusive concept than mean or
standard deviation.
 From the distribution we can deduce the peak value of frequency and
symmetry of the data range.

1. Normal Distribution / Symmetrical / Bell shaped Distribution

Normal Distribution is commonly used type. Here the values are symmetric
about the center/mean and area gives the value of probability. It Indicates
Process is Stable.

60

50

40

30

20

10

0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
6. HISTOGRAM…

2. Comb Type Distribution / Multi Modal Distribution

When class interval becomes smaller than the least count, then we will get
comb type distribution, OR, If large no. of Bins exist – It indicates
Data is rounded off.

30

25

20

15

10

0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
6. HISTOGRAM…

3. Positively Skewed Type Distribution

Mean value of the Histogram is located to the Left side from the
centre & is Asymmetrical in nature – This shape occurs when there
is 100% screening.

70

60
50

40
30

20
10

0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
6. HISTOGRAM…

4. Negatively Skewed Type Distribution

Mean value of the Histogram is located to the Right side from the
centre & is Asymmetrical in nature – This shape occurs when there
is 100% screening.

70

60
50

40

30

20
10

0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
6. HISTOGRAM…
5. Plateau Type Distribution

 The Frequency in each class is more or less same – This shape occurs
when there is Mix-Up of Lot / A mixture of several distributions having
different mean value.
 The plateau might be called a ―multimodal distribution.‖
 Several processes with normal distributions are combined.
 Because there are many peaks close together, the top of the distribution
resembles a plateau.
e.g.: Supplier mixing lots from different process

100

90

80

70

60

50

40

30

20

10

0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 100
6. HISTOGRAM…

6. Twin Peak / Bimodal Type Distribution

The frequency is low in the middle of range data and there is peak on
either side – This shape occurs when two distributions with widely
different values are mixed.

40
35
30
25
20
15
10
5
0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
6. HISTOGRAM…

7. Isolated Peak Type Distribution

There is small isolated peak in addition to general type of histogram –

This shape occurs when there is small inclusion of data from
different distribution / measurement error.

60

50

40

30

20

10

0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
6. HISTOGRAM…

8. Truncated / Heart Cut Type Distribution

 This shape is like a normal distribution with the tails cut off.
 The supplier might be producing a normal distribution of material
and then relying on inspection to separate what is within
specification limits from what is out of spec.
 The resulting shipments to the customer from inside the
specifications are the heart cut

100

90

80

70

60

50

40

30

20

10

0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 100
6. HISTOGRAM…

9. Dog Food Type Distribution

 The dog food distribution is missing something—results near the average.

 If a customer receives this kind of distribution, someone else is receiving

a heart cut, and the customer is left with the ―dog food,‖ the odds and
ends left over after the master‘s meal.

 Even though what the customer receives is within specifications, the

product falls into two clusters: one near the upper specification limit and
one near the lower specification limit.
90

80

70

60

50

40

30

20

10

0
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 100
6. HISTOGRAM…

 Skewness and Kurtosis :

 Skewness is a measure of symmetry, or more precisely, the lack of
symmetry. A distribution, or data set, is symmetric if it looks the
same to the left and right of the center point.
 Kurtosis is a measure of whether the data are peaked or flat
relative to a normal distribution. That is, data sets with high
kurtosis tend to have a distinct peak near the mean, decline rather
rapidly, and have heavy tails. Data sets with low kurtosis tend to
have a flat top near the mean rather than a sharp peak. A uniform
distribution would be the extreme case.
 The Histogram is an effective graphical technique for showing
both the skewness and kurtosis of data set.
6. HISTOGRAM…

 Skewness :For data Y1, Y2, ..., YN, the formula for skewness is:

where ‗Ybar‘ is the mean, ‗s‘ is the standard deviation, and ‗N‘ is the
number of data points.
 The skewness for a Normal Distribution is zero, and any
symmetric data should have a skewness near zero.
 Negative values for the skewness indicate data that are skewed
Left and Positive values for the skewness indicate data that are
skewed Right.
 By skewed left, we mean that the left tail is long relative to the
right tail. Or, the mass of the distribution is concentrated on the
right of the figure .
 By skewed right means that the right tail is long relative to the
left tail.Or, the mass of the distribution is concentrated on the left
of the figure.
6. HISTOGRAM…

 Kurtosis : For data Y1, Y2, ..., YN, the formula for kurtosis is:

where ‗Ybar‘ is the mean, ‗s‘ is the standard deviation, and ‗N’ is the
number of data points.
 The kurtosis for a Normal Distribution is three.
 Positive kurtosis indicates a ―Peaked" distribution and Negative
kurtosis indicates a ―Flat" distribution.
7. SCATTER DIAGRAM

Definition:
 A visual & statistical testing tool.
 Analyzes strength & relationship between 2 variables.
 Involve correlation to establish significant relationship.
 Arrive at Quantitative conclusion on relationship.

When to Use Scatter Diagram:

 In problem solving to establish a Root cause
 Examine root cause theories in C & E
 To confirm a Hypothesis
7. SCATTER DIAGRAM

SCATTER DIAGRAM :
It helps in identifying the relation between two corresponding variables e.g.
1) Hardness Vs Cooling time.
2) Hardness Vs Carbon Content.
3) Fuel consumption Vs Speed of the Car.

 Types of Correlation
 Positive Correlation
 Negative Correlation
 No Correlation.

 Demerit of Scatter Graph – The Scale makes a difference during plotting

Scatter Graph i.e. The Scatter Graph may mislead.
 To Overcome it we have to calculate Correlation Coefficient along with
Scatter Graph.
7. SCATTER DIAGRAM…

Positive Correlation Positive Correlation may Negative Correlation

be present

Negative Correlation may No Correlation Correlation Exists but

Non Linear
be present
7. SCATTER DIAGRAM…

How to Plot Scatter Graph –

1. Collect pairs of data where a relationship is suspected.

2. Draw a graph with the Independent variable on the Horizontal

axis (c) and the Dependent variable on the Vertical axis (E).
For each pair of data, put a dot or a symbol where the x-axis
value intersects the y-axis value. (If two dots fall together,
put them side by side, touching, so that you can see both.)

3. Look at the pattern of points to see if a relationship is

obvious. If the data clearly form a line or a curve, you may
stop. The variables are correlated. You may wish to use
regression or correlation analysis now.
7. SCATTER DIAGRAM…
CORRELATION COEFFICIENT (r)
 In order to understand the strength of the relation in quantitative terms, it
is useful to calculate the correlation coefficient.
 If (r) is the Quantitative Measure of Co-relation Between Two Variables.
 Two variable can be highly related to one another but Neither is Caused by
Other.
 To know the % of Dependency we use to calculate (r2 value) .

Thumb Rule POSITIVE NEGATIVE

STRONG CORELATION r ≥ 0.8 r ≤ - 0.8

MODERATE CORELATION 0.5 ≤ r < 0.8 -0.5 ≥ r > -0.8

WEAK CORELATION r < 0.5 r > -0.5

7. SCATTER DIAGRAM…

 Interpretation of ‗r‘ value :

• ‗r‘ will range from -1 to +1.
• -1 indicates very strong –ve correlation.
• +1 indicates very strong +ve correlation.
• 0 indicates no correlation.
• ‗r‘ > 1 indicates miscalculation.

 How to Calculate ‗r‘ value:

 r = S(xy) /SQRT [S(xx) * S(yy)]
• Where ‗x‘ & ‗y‘ are two related variables. ‗S‘ is called ―Co-Variance‖.
• OR, y = f(x)
• OR, x – Input & y- Output
• OR, x- cause & y- effect.
 S(xx) = Summation (xi2) – [(Summation of xi)2/n].
 S(yy) = Summation of (yi2) – [(Summation of yi)2/n].
 S(xy) = Summation of (xiyi) – [(Summation of xi)*Summation of yi) /n].
 Degree of Dependence (r2) = (r * r * 100) %.
7. SCATTER DIAGRAM…
Scatterplot of y vs x
2 2
S# x y x y xy 0.93

1 8.6 0.889 73.960 0.790 7.645 0.92

2 8.9 0.884 79.210 0.781 7.868 0.91

3 8.8 0.874 77.440 0.764 7.691 0.90

4 8.8 0.891 77.440 0.794 7.841

y
0.89
5 8.4 0.874 70.560 0.764 7.342
0.88
6 8.7 0.886 75.690 0.785 7.708
7 9.2 0.911 84.640 0.830 8.381 0.87

8 8.6 0.912 73.960 0.832 7.843 0.86

8.2 8.4 8.6 8.8 9.0 9.2 9.4
9 9.2 0.895 84.640 0.801 8.234 x

10 8.7 0.896 75.690 0.803 7.795

11 8.4 0.894 70.560 0.799 7.510 Scatterplot of y vs x
12 8.2 0.864 67.240 0.746 7.085 1.00

13 9.2 0.922 84.640 0.850 8.482

14 8.7 0.909 75.690 0.826 7.908 0.95

15 9.4 0.905 88.360 0.819 8.507

16 8.7 0.892 75.690 0.796 7.760 0.90

y
17 8.5 0.877 72.250 0.769 7.455
18 9.2 0.885 84.640 0.783 8.142 0.85
19 8.5 0.866 72.250 0.750 7.361
20 8.3 0.896 68.890 0.803 7.437
0.80
21 8.7 0.896 75.690 0.803 7.795 8.2 8.4 8.6 8.8 9.0 9.2 9.4
x
22 9.3 0.928 86.490 0.861 8.630
23 8.9 0.886 79.210 0.785 7.885
24 8.9 0.908 79.210 0.824 8.081 S(xx) = 2.87867
25 8.3 0.881 68.890 0.776 7.312
26 8.7 0.882 75.690 0.778 7.673 S(yy) = 0.00841
27 8.9 0.904 79.210 0.817 8.046
28 8.7 0.912 75.690 0.832 7.934 S(xy) = 0.09129
29 9.1 0.925 82.810 0.856 8.418
30 8.7 0.872 75.690 0.760 7.586 r= 0.5868
Total= 263.2 26.816 2312.02 23.98 235.357
r2 = 34.4285
7. SCATTER DIAGRAM…
7. SCATTER DIAGRAM…

Median method for analyzing

correlation
7. SCATTER DIAGRAM…

1. Divide points on the graph into four quadrants. If there are X points
on the graph,

2. Count X/2 points from top to bottom and draw a horizontal line.

3. Count Y/2 points from left to right and draw a vertical line.
4. If number of points is odd, draw the line through the middle point.
5. Count the points in each quadrant. Do not count points on a line.
6. Add the diagonally opposite quadrants. Find the smaller sum and the
total of points in all quadrants.
7. A = points in upper left + points in lower right
8. B = points in upper right + points in lower left
9. Q = the smaller of A and B
10. N=A+B
11. Look up the limit for N on the trend test table.
12. If Q is less than the limit, the two variables are related.
13. If Q is greater than or equal to the limit, the pattern could have
occurred from random chance
7. SCATTER DIAGRAM…
7. SCATTER DIAGRAM…

 The ZZ-400 manufacturing team suspects a relationship between

product purity (percent purity) and the amount of iron (measured
in parts per million or ppm). Purity and iron are plotted against
each other as a scatter diagram, as shown in the figure below.
 There are 24 data points. Median lines are drawn so that 12
points fall on each side for both percent purity and ppm iron.
 To test for a relationship, they calculate:
A = points in upper left + points in lower right = 9 + 9 = 18
B = points in upper right + points in lower left = 3 + 3 = 6
Q = the smaller of A and B = the smaller of 18 and 6 = 6
N = A + B = 18 + 6 = 24
 Then they look up the limit for N on the trend test table. For N =
24, the limit is 6.
 Q is equal to the limit. Therefore, the pattern could have
occurred from random chance, and no relationship is
demonstrated.
7. SCATTER DIAGRAM…
7. SCATTER DIAGRAM…

STRATIFICATION IN SCATTER DIAGRAM :

 When there is a stratification factor, you can obtain vital

information by distinguishing the data of different strata by

different colors or symbols.

 For this purpose, you should always take care to keep a

record of the data‘s origin and background which will become

a useful information for analysis afterwards .

7. SCATTER DIAGRAM…

Stratification in Scatter Diagram

When company A‘s and company B‘s data are combined

discriminately, there appears to be no correlation but when
distinguished a clear correlation becomes apparent.
7. SCATTER DIAGRAM…

Stratification in Scatter Diagram

When company A‘s and company B‘s data are combined

discriminately, there appears to be a correlation but when
distinguished the correlation disappears
7. SCATTER DIAGRAM…

Stratification in Scatter Diagram

7. SCATTER DIAGRAM…

Stratification in Scatter Diagram

7. SCATTER DIAGRAM…

Benefits of Scatter Diagram

 Trends & patterns of different measures are

tracked.

 Better process management in variable analysis.

 Relationship establishment tool.