Chapter 8

Quality Tools

Faculty of Industrial Management

Universiti Malaysia Pahang
26300 Gambang, Pahang Darul Makmur
Tel: +609 549 2166 Fax: +609 549 2167
www.fim.ump.edu.my
Learning Outcomes

✓ Understand the concept of quality tools.

✓ Use the quality tools for decision making.
✓ Analyze data & graphically present the results.
 Introduction
• Also well-known as seven QC tools.

• The tools comprised of simple

graphical & statistical techniques,
which are helpful in solving critical
quality issues.
• Focus on numerical data measure &
calculation.

• Originated from Japan when the it was

undergoing major quality revolution & had
become a mandatory topic in Japanese’s
industrial training program
Check Sheet
It is a data collection tool, to ensure that data are collected carefully,
systematically & accurately.
Check Sheet: When to Use

✓ When the data can be observed & collected repeatedly.

✓ When collecting data on frequency
✓ When identifying patterns of events, problems, defects, & defect
location.
✓ When identifying defect causes.
Check Sheet: Example
Types of non-conformity of wooden components in Company ABC.
Types of Non-conformity Tally Frequency
|||| |||| |||| |||| |||| |||| |||| |||| |||| ||||
Size out of |||| |||| |||| |||| |||| |||| |||| |||| |||| ||||
194
specification |||| |||| |||| |||| |||| |||| |||| |||| |||| ||||
|||| |||| |||| |||| |||| |||| |||| |||| ||||
Strain/rot |||| |||| |||| |||| |||| |||| | 31
|||| |||| |||| |||| |||| |||| |||| |||| |||| ||||
Fuzzy grain |||| |||| |||| |||| |||| |||| |||| |||| |||| |||| 105
||||
|||| |||| |||| |||| |||| |||| |||| |||| |||| ||||
Machine tear-out 61
|||| |||| |
Burn marks |||| |||| |||| |||| |||| |||| |||| |||| |||| 44
Raised grain |||| 4
Check Sheet is not A Check List
A tally sheet to collect A tool used to ensure all
data on frequency of steps/actions have been
occurrence taken

Often a standard form is

Custom designed by user
used

1 of 7 quality tools Not 1 of 7 quality tools

Ex: Check sheet to

document reasons for Ex: All items in case cart are
present before surgery
interruptions
Check Sheet Check List
Histogram
A bar graph that shows the distribution of data.
Histogram of Customer Complaints for Electric Stove
500 460
450
400
350
290
Frequency

300 265
250
193
200
150 135
100 53
46
50
0
Burner Drawer Front Oven door Oven Rear Timer
control rollers burners regulators burners
Parts Replacement
Histogram: When to Use
✓Summarize large data sets graphically
✓Compare measurements to specifications
✓Communicate information to the team
✓Assist in decision making
Constructing Histogram: Ungrouped Data

1. Establish an array; an arrangement of raw numerical data in

ascending or descending order.
2. Tabulate the frequency of each value (category) under the tally
column. This step can be eliminated.
3. Tabulate numerical value of frequency.
4. Construct the histogram.
Histogram-Ungrouped Data: Example
Number of Daily Accounting Tally of Number of Daily Accounting
Errors Errors
0 1 3 0 1 0 1 0 Number of Non-
Tally Frequency
conformance
1 5 4 1 2 1 2 0
0 |||| |||| 15
1 0 2 0 0 2 0 1 ||||
2 1 1 1 2 1 1 1 |||| |||| 20
0 4 1 3 1 1 1 |||| ||||
2 |||| ||| 8
1 3 4 0 0 0 0
3 |||| 5
1 3 0 1 2 2 3
4 ||| 3
5 | 1
Histogram-Ungrouped Data: Example
Number of Non-
Tally Frequency Number of Daily Accounting
conformance
Errors
0 |||| |||| 15
25
|||| 20
20
1 |||| |||| 20 15

Frequency
15
|||| |||| 8
10
2 |||| ||| 8 5
5 3
1
3 |||| 5 0
4 ||| 3 0 1 2 3 4 5
Number of non-conformance
5 | 1
Constructing Histogram: Grouped Data
Histogram-Grouped Data: Example

Determine Frequency Distributions and draw a histogram based on the

following data set.

12 18 27 31 40 42
14 20 27 32 40 51
14 20 27 32 40 56
14 21 29 34 40 60
16 23 31 36 40 65
Histogram-Grouped Data: Example
Step 1: Calculate the range of the data set.
Range = Max – Min = 65 – 12 = 53.

Step 2: Divide the range by the number of groups you want & then
round up.
Number group/class = 6.

53
𝐶𝑙𝑎𝑠𝑠 𝑊𝑖𝑑𝑡ℎ = = 8.8 → 9
6
Histogram-Grouped Data: Example
Step 3: Use the class width Step 4: Develop histogram
to create groups & find
frequency of each group
Histogram
Classes Frequency
12-20 8
21-29 6 8

30-38 6 6 6 6
39-47 6
48-56 2 2 2
57-65 2
12-20 21-29 30-38 39-47 48-56 57-65
Interpretation of Histogram

1. How well is the histogram centered? The centering of the data

provides information on the process aim about some mean or
nominal value.
2. How wide is the histogram? Looking at histogram width, which
defines the variability of the process about the aim.
3. What is the shape of the diagram? Skewness or
peakedness/flatness
 Skewness
Negatively skewed Normal: No skew Positively skewed

Skewness: The degree of symmetry in the variable distribution

Threshold: -2 ≤ skewness ≤ 2 (Curran et al., 1996; West et al., 1995; Gliselli et al.,
1981)
Kurtosis

Kurtosis: The degree peakedness/flatness in the variable distribution

Threshold: -7 ≤ Kurtosis ≤ 7 (Curran et al., 1996; West et al., 1995).
Pareto Diagram

Identified by Vilfredo Pareto (1848 to

1923), conducted studies of the
distribution of wealth.

A Pareto diagram is a graph that ranks

data classifications in descending order
from left to right.
80/20 Rule
Pareto principle was known as the 80/20
rule, which was used in materials
management for ABC analysis.

Approximately 80% of the problems are

created by approximately 20% of the
causes.
80/20 Rule
✓ 80% of quality issues are caused by 20%
of the steps in the process.

✓ 80% of total sales come from 20% of

customers.

✓ 80% of the price of products are the

effect of 20% of features.

✓ 80% of profits come from 20% of stocks.

80/20 Rule

Most effects come from relatively few

causes; that is, 80% of the effects come
from 20% of the possible causes.

The 20% of the possible causes are referred

to as the ‘vital few’; the remaining causes
are referred to as the ‘useful many’.
 Tabulate the items, their
Constructing Arrange the items in
descending order of their
contributions as well as in
percent of total &
Pareto Diagram individual contributions cumulative contribution of
the items.

Draw X & Y axes. Items are

Plot & connect points for represented on X. Y
Draw bars representing the
cumulative contributions at Represents contribution of
contributions of each item
the end of each item each item & cumulative
percent*

*Unlike other graphs, Pareto diagram has two Y-axes.

One on the left representing numbers & the one on right representing % contributions
Pareto Diagram: Example 1
Construct a Pareto diagram for replacement parts for an electric stove.
Six month data are: oven door, 193; timer, 53; front burners, 460; rear
burners, 290; burner control, 135; drawer rollers, 46; & oven
regulators, 265.
Pareto Diagram: Example 1

Parts
Frequency Cumulative Frequency Cumulative %
Replacement
Front burners 460 460 31.90%
Rear burners 290 750 52.01%
Oven regulators 265 1015 70.39%
Oven door 193 1208 83.77%
Burner control 135 1343 93.13%
Timer 53 1396 96.81%
Drawer rollers 46 1442 100.00%
 Pareto diagram for replacement parts for an electric
stove
120%
1600
1400 96.81% 100.00%100%
93.13%
1200 83.77% 80%

Cumulative %
Frequency

1000 70.39%
800 60%
52.01%
600 40%
460
400 31.90%
290 265
193 20%
200 135
53 46
0 0%
Front burners Rear burners Oven Oven door Burner Timer Drawer rollers
regulators control
Parts Replacement
Pareto Diagram
Paint nonconformities for a one month period for a
riding lawn mower manufacturer are: blister, 212;
light spray, 582; drips, 227; overspray, 109; splatter,
141; bad paint, 126; runs, 434; and other, 50. You
are required to:
1. Construct a Pareto diagram.
2. Suggest areas for improvement.
Pareto Diagram: Exercise A1
It is evident from
the Pareto chart
that 80% of the
paint non
conformities come
from light spray
while least source
is from others
category
Cause & Effect (C&E) Diagram
A picture composed of lines & symbols to represent the relationship
between an effect & its causes.

C&E diagrams are used to investigate:

1. A “bad” effect & to take action to correct the causes.
2. A “good” effect & to learn those causes that are responsible.
 #1. Clearly identify the effect.
Problem is placed in a box at the
Constructing C&E
end of a line Diagram
#2. Identify the causes.
#5. Analyze the diagram & take
Brainstorming is the usual method corrective action
for identifying these causes.

#4. Draw the effect box & the

#3. Build the diagram.
center line.
Organize the causes & sub-causes
Connect the potential causes
in diagram format.
boxes to the center line.
Constructing C&E Diagram
Causes are usually broken down into the major causes (e.g., work
methods, materials, machine, man/people & environment, etc.)
Each major cause is further subdivided into numerous minor causes.
C&E Diagram: Example

Construct C&E diagrams to identify the potential root-causes of:

a) Lack of responsiveness to customer.
b) Road congestion.
c) Traffic accidents in the UMP main junction.
d) Poor performance in a mid-term examination.
C&E Diagram for lack of
responsiveness to customer
Benefits of C&E Diagram

Helps determine root causes

Indicates possible causes Analyze actual conditions
of variation for improvement
Educates & trains personnel in
decision making & corrective Encourage group participation
action
FISH BONE DIAGRAM: IA@3
Scatter Diagram
A visualization of the relationship
between two variables measured on the
same set of individuals.

It shows how much one variable is

affected by another.

The relationship between two variables is

called correlation.
Constructing Scatter Diagram

1. Collect 2 sets of data & create a summary table of the data.

2. Draw a diagram labelling the horizontal & vertical axes.
The “cause” variable (IV) is labelled on the X axis. The “effect”
variable (DV) is labelled on the Y axis.
3. Plot the data pairs on the diagram.
4. Interpret the scatter diagram for direction & strength.
Correlation Patterns in Scatter Diagram

No correlation Strong positive linear Strong negative linear

correlation correlation
Correlation Patterns in Scatter Diagram

Weak positive linear Weak negative linear Non-linear correlation

correlation correlation
Scatter Diagram: Mathematical Calculation
Calculate the gradient (m) of the relationship between X & Y:
𝑛 σ 𝑥𝑦 − σ 𝑥 σ 𝑦
𝑚=
𝑛 σ 𝑥2 − σ 𝑥 2

Determine the equation of the line of best fit (or trend line) that best
represents the data on a scatter plot.
Calculate intercept: 𝑏 = 𝑦ത − 𝑚𝑥ҧ
Line of best fit: 𝑦 = 𝑚𝑥 + 𝑏
Draw the equation of the line of best fit on the scatter plot.
Correlation Coefficient (r)
It describes the goodness of fit of the linear model.

𝑛(σ 𝑥𝑦) − (σ 𝑥)(σ 𝑦)

𝑟=
𝑛 σ 𝑥2 − σ 𝑥 2 𝑛 σ 𝑦2 − σ 𝑦 2

Range: -1 ≤ r ≤ 1

Positive & negative signs tell whether there is a positive/negative correlation.

The closer the r to 1, the better the fit, with a value of 1 meaning that all points fall
on the line.
Scatter Diagram: Example
The following table shows the number of pages Number of Weight
& the weight (in gram) for each 10 books: Observation
page (X) (gram) (Y)
1. Construct a scatter diagram of the
relationship between X & Y. 1 80 160
2. Calculate the gradient (m) of the relationship 2 130 270
between X & Y. 3 100 180
3. Determine and draw the equation of the line 4 140 290
of best fit that best represents the data on a 5 115 230
scatter plot. 6 90 180
4. Calculate the correlation coefficient (r) of the 7 160 320
association between X & Y. 8 140 270
5. Based on the scatter plot, describe what is 9 105 210
the association between X & Y. 10 150 300
Scatter Diagram: Example
Scatter Diagram
350

300

250
Weight (gram)

200

150

100

50

0
0 20 40 60 80 100 120 140 160 180
Number of pages
Scatter Diagram: Example
Obs. X Y X*Y X^2 Y^2

Calculate the gradient (m) of the 1 80 160 12,800 6,400 25,600

relationship between X and Y 2 130 270 35,100 16,900 72,900

3 100 180
18,000 10,000 32,400
𝑛 σ 𝑥𝑦 − σ 𝑥 σ 𝑦 4 140 290 40,600 19,600 84,100
𝑚=
𝑛 σ 𝑥2 − σ 𝑥 2 5 115 230 26,450 13,225 52,900
6 90 180 16,200 8,100 32,400
7 160 320 51,200 25,600 102,400
Calculate intercept: 𝑏 = 𝑦ത − 𝑚𝑥ҧ 8 140 270 37,800 19,600 72,900

Line of best fit: 𝑦 = 𝑚𝑥 + 𝑏 9 105 210 22,050 11,025 44,100

10 150 300 45,000 22,500 90,000

Sum 1210 2410 305,200 152,950 609,700

Average 121 241
Scatter Diagram: Example
Obs. X Y X*Y X^2 Y^2
1 80 160 12,800 6,400 25,600
𝑛 σ 𝑥𝑦 − σ 𝑥 σ 𝑦 2 130 270 35,100 16,900 72,900
𝑚= = 2.078
𝑛 σ 𝑥2 − σ 𝑥 2 3 100 180 18,000 10,000 32,400
4 140 290 40,600 19,600 84,100
5 115 230 26,450 13,225 52,900
Calculate intercept (the expected 6 90 180 16,200 8,100 32,400
mean value of Y when all X=0):
7 160 320 51,200 25,600 102,400
𝑏 = 𝑦ത − 𝑚𝑥ҧ = −10.438
8 140 270 37,800 19,600 72,900
9 105 210 22,050 11,025 44,100
Best fit equation: 10 150 300 45,000 22,500 90,000
𝑦 = 𝑚𝑥 + 𝑏 = 2.078 𝑥 − 10.438 Sum 1210 2410 305,200 152,950 609,700
Average 121 241
Scatter Diagram: Example
350
y = 2.078x - 10.436
300

250

Weight (gram)
Draw the 200
equation of the
150
line of best fit.
100

50

0
0 20 40 60 80 100 120 140 160 180
Number of pages
Scatter Diagram: Example
Obs. X Y X*Y X^2 Y^2
Calculate the correlation coefficient (r) of
1 80 160 12,800 6,400 25,600
the association between X and Y:
2 130 270 35,100 16,900 72,900

𝑛(σ 𝑥𝑦) − (σ 𝑥)(σ 𝑦) 3 100 180 18,000 10,000 32,400

𝑟= 4 140 290 40,600 19,600 84,100
𝑛 σ 𝑥2 − σ𝑥 2 𝑛 σ 𝑦2 − σ𝑦 2
5 115 230 26,450 13,225 52,900
6 90 180 16,200 8,100 32,400
r = 0.989
7 160 320 51,200 25,600 102,400
8 140 270 37,800 19,600 72,900
What is the association between X and Y? 9 105 210 22,050 11,025 44,100

Positive and strong correlation; the more 10 150 300 45,000 22,500 90,000
the number of pages, the heavier the Sum 1210 2410 305,200 152,950 609,700
book. Average 121 241
Control Chart
A graph used to study how a process changes over time.

Data are plotted in time order.

A control chart always has a central line for the average, an upper line
for the upper control limit and a lower line for the lower control limit.

These lines are determined from historical data.

Structure of Control Chart
Conditions (States) of Process

Process is stable or Process is unstable or

“in control” “out of control”

Plotted points are within the control limits Plotted points are outside the limits OR follow
AND follow a random pattern (random a non-random pattern inside the control limits.
causes) An assignable cause is present
Objectives of Control Chart
✓ Identify whether the process variation is consistent (in control) or
unpredictable (out of control, affected by special causes of variation).
✓ Obtain information about quality specification.
✓ Provide a visual baseline for continuing process improvement over
time
✓ Monitor & control machines parameter (e.g., speed, rpm, pressure)
& processes.
✓ Supply information (i.e., mean of process, process capability) to
customers of conformance to specification.
✓ Helps in decision making/problem solving.
2 Types of Data
• Measurable, data derived from incremental
measurements, recorded in numerical data,
Variable data usually continuous measurement.
• Examples: length, voltage, viscosity, weight,
time.

• Character, evaluated with a discrete choice,

often collected in the final inspection when
Attribute data assembled products are tested/inspected
virtually.
• Examples: good/bad, yes/no, accept/reject.
Variable Control Chart
✓ Uses amount of dispersion in a sample
(Xmax – Xmin)
✓ Uses average of a sample (sub-group). ✓ Tracks the dispersion of sample
✓ Tracks the control tendency (shows measured data in a quality
any changes in the mean value of the
process). characteristic (shows any changes in
the dispersion of the process).
✓ Graphs the variation of the
measurements in each sample over
𝑋ത Chart time R Chart
Control chart for Control chart for
mean range
Important Notes
𝑋ത and R charts are used for variable data when the sample size of the
sub-group is 2-5.
When the subgroup size is larger, s Charts are used instead of R charts
where s is the standard deviation of the sub-group.
Symbols & Terminologies in Control Chart
Symbols Description
m Number of subgroup
n number of observation in each sample (size of subgroup
sample)
xi Value (measurement) of an individual item i.
𝑥1 +𝑥2 + 𝑥3 + 𝑥4 +⋯+𝑥𝑛
𝑥ҧ Sample mean; 𝑥ҧ =
𝑛
𝑋ത Mean of the sample mean; 𝑋 = ത σ 𝑥𝑖
𝑚
R Sample range; 𝑅 = 𝑥𝑚𝑎𝑥 − 𝑥𝑚𝑖𝑛
𝑅ത ത
Average range; 𝑅 =
𝑅1 + 𝑅1 + 𝑅1 +⋯+ 𝑅𝑚
𝑚
2 Methods of Constructing Control Charts
1st Method

Trial control limits for the charts are established at ± 3σ from

the central value (theoretical).

𝑋ത Chart R Chart

UCLx = 𝑋ത + UCLR = 𝑅ത + 3σR

3σx LCLx = LCLR = 𝑅ത - 3σR
𝑋ത - 3σx
2 Methods of Constructing Control Charts
2nd Method
Simplified method by using table of constants.

𝑋ത Chart 𝑅 Chart

𝑈𝐶𝐿𝑥 = 𝑋ധ + 𝐴2 𝑅ത 𝑈𝐶𝐿𝑅 = 𝐷4 𝑅ത
𝐶𝐿𝑥 = 𝑋ധ 𝐶𝐿𝑅 = 𝑅ത
𝐿𝐶𝐿𝑥 = 𝑋ധ − 𝐴2 𝑅ത 𝐿𝐶𝐿𝑅 = 𝐷3 𝑅ത

A2, D3 & D4 are factors that vary with the sub-group’s sample size.
A2, D3 & D4 are taken from Table of Constants for Control Chart.
Example
Observation
Sample
1 2 3 4 5
1 10.682 10.689 10.776 10.798 10.714
2 10.787 10.86 10.601 10.746 10.779
3 10.78 10.667 10.838 10.785 10.723
4 10.591 10.727 10.812 10.775 10.73 a) Calculate the mean of 𝑥ҧ and R.
5 10.693 10.708 10.79 10.758 10.671 b) Determine the control limits for 𝑥ҧ chart
6 10.749 10.714 10.738 10.719 10.606
7 10.791 10.713 10.689 10.877 10.603
and R chart.
8 10.744 10.779 10.11 10.737 10.75 c) Sketch 𝑥ҧ chart and R chart.
9 10.769 10.773 10.641 10.644 10.725
d) Identify if there are any data out-of-
10 10.718 10.671 10.708 10.85 10.712
11 10.787 10.821 10.764 10.658 10.708 control.
12 10.622 10.802 10.818 10.872 10.727
13 10.657 10.822 10.893 10.544 10.75
14 10.806 10.749 10.859 10.801 10.701
15 10.66 10.681 10.644 10.747 10.728
Example
Sampl Observation
𝑋ത R
a) Calculate the mean of 𝑥ҧ and R. e 1 2 3 4 5
𝐶𝐿𝑥 = 𝑋ധ = 10.782 1 10.682 10.689 10.776 10.798 10.714 10.732 0.116
𝐶𝐿𝑅 = 𝑅ത = 0.220 2 10.787 10.86 10.601 10.746 10.779 10.755 0.259
3 10.78 10.667 10.838 10.785 10.723 10.759 0.171
4 10.591 10.727 10.812 10.775 10.73 10.727 0.221
b) Determine the control limits for 𝑥ҧ chart & R chart.
5 10.693 10.708 10.79 10.758 10.671 10.724 0.119
𝑈𝐶𝐿𝑥 = 𝑋ധ + 𝐴2 𝑅ത = 10.728 + 0.58 x 0.22 = 10.856 6 10.749 10.714 10.738 10.719 10.606 10.705 0.143
𝐿𝐶𝐿𝑥 = 𝑋ധ − 𝐴2 𝑅ത = 10.605 7 10.791 10.713 10.689 10.877 10.603 10.735 0.274
8 10.744 10.779 10.11 10.737 10.75 10.624 0.669
𝑈𝐶𝐿𝑅 = 𝐷4 𝑅ത = 2.114 x 0.220 = 0.465 9 10.769 10.773 10.641 10.644 10.725 10.710 0.132
𝐿𝐶𝐿𝑅 = 𝐷3 𝑅ത = 0 10 10.718 10.671 10.708 10.85 10.712 10.732 0.179
11 10.787 10.821 10.764 10.658 10.708 10.748 0.163
12 10.622 10.802 10.818 10.872 10.727 10.768 0.25
13 10.657 10.822 10.893 10.544 10.75 10.733 0.349
14 10.806 10.749 10.859 10.801 10.701 10.783 0.158
15 10.66 10.681 10.644 10.747 10.728 10.692 0.103
Average 10.728 0.220
Example
c) Sketch 𝑥ҧ chart and R chart
X-bar Chart R Chart
10.900 0.8
10.850 0.7 0.669

10.783
10.800 10.768 0.6
10.75510.759 10.748
10.732 10.735 10.732 10.733
10.750 10.72710.724
10.705 10.710
10.692
0.5
10.700
0.4 0.349
10.650 10.624
0.274
0.3 0.259 0.25
10.600 0.221
0.171 0.179 0.163
10.550 0.2 0.143 0.132
0.158
0.116 0.119 0.103
10.500 0.1
10.450 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
X-bar UCLx CL LCLx Range UCLR CL LCLR
 Faculty of Industrial Management
Universiti Malaysia Pahang
26300 Gambang, Pahang Darul Makmur
Tel: +609 549 2166 Fax: +609 549 2167
www.fim.ump.edu.my