ID document: e.g.

PPT/L-TCIO-1

Chapter II – Reminder
of statistics &
Introduction to
Statistical Process
Operational quality and Lean
management - S7 EENG 4 Control Title of Lesson
Title of Lesson 1
ID document: e.g. PPT/L-TCIO-1

Chapter summary
1 - Basic tools of statistics (Calculation of mean value, standard deviation, quartile,
median, mode)
- Notion of population and sample
- Positional characteristics
2 - Graphical representation (Histogram, box plot, Pareto chart)
3 - Normal law (normal distribution)
- Introduction to the normal law
- The parameters of the normal law
- Construction and analysis of the histogram
4 - Use and comparison of these three indicators: Mean value, mode and median

Title of Lesson 2
ID document: e.g. PPT/L-TCIO-1

Targets of today lecture :

● Understand why statistics are important for quality management

● Knowledge of representations of the dispersion of a series of values

● Estimated standard deviation (S) usage

● Knowledge of Normal law parameters

Title of Lesson 3
ID document: e.g. PPT/L-TCIO-1

1 – Basic tools of statistics

Title of Lesson 4
ID document: e.g. PPT/L-TCIO-1

1 – Basic tools of statistics

A – Notion of population and

sample

Title of Lesson 5
ID document: e.g. PPT/L-TCIO-1

The link between statistics and Quality

a population : N pieces a sample : n pieces

Mass production : too costly to check each piece

Statistics make possible to know what’s the whole batch
characteristics by sampling inspection Title of Lesson 6
ID document: e.g. PPT/L-TCIO-1

Notion of population and sample

● In statistics, we must distinguish a population and a sample. For example, to have an
idea of the distribution of shoe sizes, we can make a study on 50 people. In this case,
we talk about a sample.

● We can theoretically multiply this measure to infinity in order to have the true value
of the average. We will then speak about population.

● Notations :
- n : size of the sample, composed of n data
- xi : value of the ith data of the sample
- N : size of the population

Title of Lesson 7
ID document: e.g. PPT/L-TCIO-1

1 – Basic tools of statistics

- Average / Mean value

- Median
- Quartile
B - Positional characteristics - Mode

Title of Lesson 9
ID document: e.g. PPT/L-TCIO-1

Nuts sampling
Si
ze
Thickness

Let’s measure few pieces : size and thickness

1rst conclusion: despite nuts are coming from the same

batch, they don’t have the same dimension …
2nd conclusion: dimensions keep in a narrow range
How to have a better description of the sampling?
Title of Lesson 10
ID document: e.g. PPT/L-TCIO-1

Measures from the sample of 30 nuts

Title of Lesson 11
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Mean value

These characteristics give an idea of the order of magnitude of the values making up
the series, as well as the position where the values in the series appear to be
located.

● The arithmetic mean :

For xi given values, the arithmetic mean is defined by :

Title of Lesson 12
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Mean value

● Mean value properties:

- Easy to calculate
- Sensitive to extreme values, it is sometimes necessary to remove aberrant values.

● Let's go back to our example of nuts :

Size Mean value : = 12,85 mm

Thickness Mean value : = 6,27 mm

Title of Lesson 13
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Median

● Median :

The median divides the data set into 2 equal parts. There are 50% of the values on each
side of the median. It is usually represented by the symbol Me.

If xi is the ith data of the sample ordered in increasing order of size n, then the median is
defined by :

Title of Lesson 14
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Median

● Median properties :

- Fast calculation of the median

- Not influenced by extreme/aberrant values 15 values
- Influenced by the number of data

Me =
(12,86+12,86)/2
15 values

Title of Lesson 15
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Quartile

● Quartile :

In descriptive statistics, a quartile is each of the three values that divide the sorted
data into four equal parts, so that each part represents 1/4 of the population sample.

Following this logic: the second quartile is therefore the median of the sample.

Title of Lesson 16
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Quartile

● The first quartile (Q1) is the smallest value in the series such that at least 25% of the
values are less than or equal to Q1.

● The third quartile (Q3) is the smallest value in the series such that at least 75% of the
values are less than or equal to Q3.

Title of Lesson 17
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Quartile

25% of 30 = 7,5
1rst Quartile is place 8 = 12,79mm
8 values

23 values Me
75% of 30 = 22,5
3rd Quartile is place 23 = 12,87mm

Title of Lesson 18
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Mode

● Mode:

The mode also characterizes the position of the distribution. The mode is the value
where the frequency is the most important. It will therefore correspond to the
largest bar on the histogram.

● Mode properties:

- The mode does not always exist, and when it does, it is not always unique.
- We will take up this notion of mode after the definition of graphical representations
and the definition of classes.

Title of Lesson 19
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Estimated standard

deviation (S)
● L’écart-type estimé :

The estimation of the standard deviation of a population is done on a sample. The real
standard deviation (σ) is calculated using the values of the complete population. S (or
σn-1) is therefore the best estimation of σ on a sample of this population:

To calculate the estimated standard deviation we need to know :

: the arithmetic mean (mean value) of the sample

n : the sample size
Title of Lesson 20
ID document: e.g. PPT/L-TCIO-1

Let’s dig into this formula

X10 𝑿= 12,85 mm
X 22

Nut size

For X10 : Xi- 𝑿 = -0,06mm : distance to mean value

2
( 𝑥𝑖 − 𝑋 ) = (distance)² to mean value => makes it >0
2
The larger the distance is, the bigger( 𝑥 𝑖 − 𝑋 ) is
2
For X10 : ( 𝑥 𝑖 − 𝑋) =0,0036
X 9 while distance to mean value is
2
For X22 : ( 𝑥 𝑖 − 𝑋)
only X 3 between X10 and X22
=0,0004
Title of Lesson 21
ID document: e.g. PPT/L-TCIO-1

Same mean value but standard deviation different

𝑿
Big standard deviation
Xi

Xi 𝑿 small standard deviation

Title of Lesson 22
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Standard deviation (σ)

● Standard deviation :

The true value of the standard deviation, calculated on a whole population, is expressed
by :

● In this case, we need to know :

- m: the actual average over the population

- N : the size of the population

Title of Lesson 23
ID document: e.g. PPT/L-TCIO-1

Positional characteristics – Standard deviation (σ)

● Standard deviation properties :

- The standard deviation characterizes the dispersion of a series of values. The

smaller the standard deviation, the more the values are grouped around the
arithmetic mean and the more homogeneous the population is. .

- In the case where the distribution follows a normal distribution, the standard
deviation allows us to find the percentage of the population belonging to an
interval centered on the mathematical expectation.

- The standard deviation takes into account all the data, it is the best dispersion
characteristic.

Title of Lesson 24
ID document: e.g. PPT/L-TCIO-1

2 – Graphical representation

Title of Lesson 25
ID document: e.g. PPT/L-TCIO-1

2 – Graphical representation

- Definition

- Construction
A - Histogram

Title of Lesson 26
ID document: e.g. PPT/L-TCIO-1

Graphical representation - Histogram

● Definition :

A histogram is a contiguous bar graph (rectangle) whose areas are proportional to the
frequencies.

A histogram graphically shows the following points:

- The centering of the data

- An idea of the overall dispersion of the data
- The symmetry (or not) of the distribution
- The possible presence of aberrant values
- The possible presence of multiple modes (Reminder: a mode is the place where the
value(s) are the most represented in terms of frequency).
Title of Lesson 27
ID document: e.g. PPT/L-TCIO-1

Graphical representation - Histogram

● Construction of the histogram in 4 steps :

- Step 1: The choice of the number of classes (Kt)

The number of classes (rounded up to the next integer) is given by the following
relationship:

That we can approximate to :

- N : number of values

The number of classes is generally limited to 20 classes. For the nuts size example : √30 ~ 5,5
rounded to 6

Title of Lesson 28
ID document: e.g. PPT/L-TCIO-1

Graphical representation - Histogram

- Step 2: Calculation of the class interval (Ht)

The first step is to calculate the measurement range (Wt) of the sample:
Wt = For the nuts size example :
Wt = 12,92-12,77 = 0,15
We can then calculate the class interval (Ht):
Ht : 0,15/6= 0,025
Caliper resolution = 0,01
Ht rounded to : 0,03

- The theoretical class interval should be rounded to a multiple of the measurement resolution of the
instrument. The measurement resolution depends on the instrument used.

- If the class interval is not a multiple of the instrument's resolution, each class will not contain the same
measurement deviation.

Title of Lesson 29
ID document: e.g. PPT/L-TCIO-1

Graphical representation - Histogram

- Step 3: Calculation of the lower limit value

The lower limit value is equal to the smallest value minus half the resolution.

● Example :
For the nuts size example :
Lower value = 12,77mm
We have :
Caliper resolution = 0,01
- Lower value = 9,3
- Measurement resolution = 0,1 Lower limit value = 12,77-0,005=12,765

Lower limite value = 9,3 – 0,05 = 9,25

Title of Lesson 31
ID document: e.g. PPT/L-TCIO-1

Graphical representation - Histogram

- Step 4: Construction of the histogram

The construction of the histogram is done :

- by counting the number of values in each class

- by representing a rectangle with height proportional to the number of values and

width proportional to the class interval.
For the nuts size example : Mode

Title of Lesson 32
ID document: e.g. PPT/L-TCIO-1

2 – Graphical representation

B - Box plot

Title of Lesson 33
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Box plot

The box plot is commonly used to represent the distribution of populations.

● The box:

- Its height corresponds to 50% of the population.

- The bottom of the box corresponds to the first quartile (Q1: 25% of the population)
and the top to the third quartile (Q3: 75% of the population).

The height of the box corresponds to the interquartile distance (Q3 - Q1 = 50% of the
population).

Title of Lesson 34
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Box plot

● Whiskers (Moustaches) :

These are the lines that run on either side of the box. They represent the range of the
data, if there are no outliers (aberrant values) :

To define the outliers, we apply Tukey's Law by calculating :

Low limit value =

Upper limit value =

Outliers are the points outside the high and low boundary values defined by Tukey's rule.
They are represented by asterisks (*).

Title of Lesson 35
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Box plot

● Separation line:
The line separating the box in 2 is the median (50% of values above and 50% below).

This representation is particularly

useful to visualize the symmetry of a
distribution or to compare several
distributions.

Note: There is no standardization of

the representation of the plot boxes.
* One can therefore find many variants
in its representation.

Title of Lesson 36
ID document: e.g. PPT/L-TCIO-1

Example of the nuts size

Gives a better view than the quartile position
we had at the begining of this lecture

3rd Quartile
A big part of the measures
are at 12,86mm
Mediane

1rst Quartile

Title of Lesson 37
ID document: e.g. PPT/L-TCIO-1

2 – Graphical representation

C – Pareto analysis

Title of Lesson 38
ID document: e.g. PPT/L-TCIO-1

Context
Very likely, in your professional life you’ll have more topics to work on than
your availability or your essources will allow you .

The question will then be : what topics should I select to work on in order
to have the better overall impact

Parerto makes it obvious

Title of Lesson 39
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

● Pareto analysis:

The purpose of the Pareto analysis is to allow to classify different elements that are
sources of trouble, of problems, by order of importance and to suggest objectively
a choice.

80% of the turnover is made by 20% of the products

80% of non-quality is due to 20% of the components
80% of breakdowns or stoppages are caused by 20% of the machines

The diagram that is drawn is also called DIAGRAM 20/80.

Title of Lesson 40
ID document: e.g. PPT/L-TCIO-1

Example of Pareto for failure symptom of a kitchen

appliance

Title of Lesson 41
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

● Construction of the Pareto chart :

- Step 1: Fill in the Pareto ranking table

The first step is to reproduce a first table (next slide) and to fill it with the different
elements to be analyzed (the different families of components). It will allow us to
classify these families and to include them in the Pareto recovery table.

Title of Lesson 42
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

- Step 1: Fill in the Pareto ranking
table

Here you have to fill in the rows of the

table with information:
- The elements (families): it can be
tools, causes, ... As an example : spare parts (SP) used to repair
- Its quantity coffee machines during year 2022

- Its percentage in relation to the total

- Its ranking in relation to the total

Title of Lesson 43
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

- Step 2: Filling in the Pareto recovery
table

On each line, you must indicate:

- The total value of the family

- Its global percentage
- The cumulative value line by line
- The global percentage accumulated
line by line

Title of Lesson 44
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

- Step 2: Filling in the Pareto recovery
table

For this step, we need to take the values

from the Pareto table on the previous
page and list them in descending
order in this new table. The families
are therefore represented from the
most important (in terms of value) to
the least important.

Note : The last line should therefore theoretically

indicate the global total of the components
being analyzed, and a cumulative percentage
of 100%.
Title of Lesson 45
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

- Step 3 : Construction of the chart

For this diagram, a double scale must be

drawn:
- On one side, the number of components
- On the other side, the percentage (from 0 to 100%
of the total number of components)

The different families are represented in

descending order of scale on the graph :

- by bars (rectangles) proportional to their total

number of components
- by dots corresponding to the cumulative total
indicated in the table taken from Pareto
Title of Lesson 46
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

- Step 4 : Pareto analysis, ABC analysis

We can now check if the distribution of the components follows the Pareto
law or not.

For that, it is enough to :

- Draw a horizontal line at 80% of the cumulated percentage, and to check if

that corresponds or not to 20% of the families (of the components, stocks,
losses...).
- The Pareto law is verified if at least 20% of the families correspond to 80%
of the cumulative percentage.

Title of Lesson 47
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

- Step 4 : Pareto analysis, ABC analysis Class C
Class B
Class A
ABC Analysis :

● Class A: products in this class generally represent 80% of the total value of
the inventory and 20% of the total number of items. It is on this point that the
ABC classification method is the heir of the Pareto principle;

● Class B: the articles generally represent 15 % of the total value of stock and
30 % of the total number of articles;

● Class C: items generally represent 5% of the total stock value and 50% of the
total number of items.

Title of Lesson 48
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

We can conclude :

● Class A must be followed with maximum attention and rigor because it is

the one that keeps the company going;

● Class B must be followed with a certain amount of concern because it

represents the star products either in a maturation phase or in a decline
phase. It is therefore necessary to clarify the products of this class in order to
adopt the right solution;

● Class C should be followed with concern as it represents products in the

launch or decline phase. For products in the launch phase, we must try to
ensure their maturation.

Title of Lesson 49
ID document: e.g. PPT/L-TCIO-1

Graphical representation – Pareto

We can then physically plot this
analysis on the Pareto chart.

The different families represented

on the x-axis are thus classified
in their category, and we can
define the actions to be taken.

Title of Lesson 50
ID document: e.g. PPT/L-TCIO-1

3 – Normal law

Title of Lesson 51
ID document: e.g. PPT/L-TCIO-1

3 – Normal law

- Distribution

- Normal law curve

A – Introduction to normal
law

Title of Lesson 52
ID document: e.g. PPT/L-TCIO-1

Galton board experience

https://www.facebook.com/TrustMyScience/videos/planche-de-galton/
971920232964990/

391 Planche de Galton.mp4

Title of Lesson 53
ID document: e.g. PPT/L-TCIO-1

Galton board experience

Marbles drop from the top to the channels and find in the way some nails

Why do the marbles accumulates in the channels like the red curve ?
Title of Lesson 54
ID document: e.g. PPT/L-TCIO-1

Galton board

bell-shaped curve
Also called Gauss Curve
Title of Lesson 55
ID document: e.g. PPT/L-TCIO-1

Introduction to normal law

In statistics, this bell-shaped curve is called the normal law curve or Gauss curve
(named after the mathematician who studied this phenomenon).

This observation is not a coincidence, we find this distribution very often in

nature.

Example:
If we represent the histogram of the sizes of a group of 50 men, the distribution
will not be random. Many people are around 1.75m tall, and few are 1.95m
tall. Similarly, men who are 1.55m tall are quite rare. The size distribution will
therefore follow a normal distribution.

Title of Lesson 57
ID document: e.g. PPT/L-TCIO-1

Introduction to normal law

We find on the right side the distribution curve of Normal law curve
the size of the men on a sample.

In production, the distribution of the different

pieces, resulting from a machine whose
process is controlled, follows a normal law.
Indeed, this distribution is the consequence of
the central limit theorem which says:

Any system, subjected to many factors, independent of each other,

and of equivalent order of magnitude of effect, generates a
distribution that follows a normal distribution.

Title of Lesson 58
ID document: e.g. PPT/L-TCIO-1

Introduction to normal law

The average position of the pieces gives a good indication of the machine's
setting position.

The width of the curve (dispersion) will give a good indication of the machine's
ability to produce pieces within a given tolerance interval.

Title of Lesson 59
ID document: e.g. PPT/L-TCIO-1

3 – Normal law

-Range

-Standard deviation
B – Normal law parameters

Title of Lesson 60
ID document: e.g. PPT/L-TCIO-1

Normal law parameters

To characterize the width of a bell curve, there are mainly 2 methods.
The first one is to calculate the range, the second one is to calculate the
standard deviation.

- The range is easy to calculate, but very imprecise when the number of
values is high (>10). Indeed, the range only takes into account the 2 extreme
values. An outlier therefore has a lot of influence on the range.

- On the other hand, the standard deviation is more precise, it is calculated

from all the values, but requires the use of a calculator. An outlier has little
influence on the standard deviation.

Title of Lesson 61
ID document: e.g. PPT/L-TCIO-1

Normal law parameters

The standard deviation S is the parameter Inflexion point

that best expresses the dispersion of

values around the mean. It is a
characteristic parameter of the normal
law.

It is possible to know very precisely the

different percentages of pieces which
are included between two values
expressed in number of standard
deviation.

Title of Lesson 62
ID document: e.g. PPT/L-TCIO-1

Normal law parameters

- 68.26 % of the values are distributed
Inflexion point
around the mean between ± 1 standard
deviation

- 95.44 % of the values are distributed

around the mean between ± 2 standard
deviation

- 99.73 % of the values are distributed

around the mean between ± 3 standard
deviation

The last result is important because it

represents the dispersion
Title of Lesson 63
ID document: e.g. PPT/L-TCIO-1

Normal law parameters

By definition, the dispersion for a normal distribution is equal to the difference

between the mean + 3 times the standard deviation and the mean - 3 times
the standard deviation. This corresponds to 99.73% of the values.

It is important to make the difference between the range and the dispersion.

Title of Lesson 64
ID document: e.g. PPT/L-TCIO-1

Normal law parameters

- The range is equal to the difference
between the largest and smallest
parts. It is measured on the histogram.

- The dispersion is equal to 6 times the

standard deviation. It is measured on
the curve of the normal law.

The range is almost always less than the Range

dispersion. Dispersion

Title of Lesson 65
ID document: e.g. PPT/L-TCIO-1

3 – Normal law

- Example of packages

C – Histogram construction

Title of Lesson 66
ID document: e.g. PPT/L-TCIO-1

Histogram construction
This part will be based on an example : parcels.

In order to establish new scales by delivered weight, a transport company has taken
the weights in kg of 92 shipments controlled with a digital scale.

The document
The table of the different weights is on Moodle. for the example
is on Moodle

We will be able to construct the histogram in several steps.

Title of Lesson 67
ID document: e.g. PPT/L-TCIO-1

Histogram construction
- Step 1: Calculation of the range

In each column the maximum and minimum values have been marked. The
sample size is n = 92 parcels.

The range of the sample is :

Title of Lesson 68
ID document: e.g. PPT/L-TCIO-1

Histogram construction
- Step 2: Number of classes
Number of classes :
On the range, all the values are distributed in
intervals of constant width called classes. The
distribution by classes is more meaningful than
the one that would consist in marking each
value with a cross.

In order to lose as little information as possible, a

rule of thumb is to choose a number of classes
close to the square root of the number of
values.

Title of Lesson 69
ID document: e.g. PPT/L-TCIO-1

Histogram construction
- Step 3: Width between classes

The width of the classes is given by the range, divided by the number of classes.

The center of the class should be a simple number that makes the calculations easier.
It doesn't matter what the ends of the classes are, they will be of little use for the
calculation.
To simplify: Ht = 5 (from - 2.5 to + 2.5 around the class
center)
The class centers are: 35, 40, 45, 50 ...
Calculation is for you

Title of Lesson 70
ID document: e.g. PPT/L-TCIO-1

Histogram construction
- Step 4: Construction of the histogram

The values are taken one by one, marked

with a cross and ticked on the table after
each transfer to the graph (on the center
of classes defined before). When
performing this operation, it is possible
that some values fall just on the class
limit: a convention suggests "right
priority".

Title of Lesson 71
ID document: e.g. PPT/L-TCIO-1

Histogram construction
- Step 4: Construction of the histogram

It is therefore necessary to transfer marks

(here crosses) with

- A constant distance between each of

them on the centers of classes.

- We will then find thanks to this step the

height of the bars of the final histogram.

Title of Lesson 72
ID document: e.g. PPT/L-TCIO-1

3 – Normal law

-Comparison parameters

- Example : packages (again)

D – Histogram analysis

Title of Lesson 73
ID document: e.g. PPT/L-TCIO-1

Histogram analysis - Parameters

Simplified parameters of comparison :

- Absolute frequency: the number of individuals in each class

- The relative frequency: it is the absolute frequency divided by the total

number of individuals

- The arithmetic mean: it is calculated by the ratio of the sum of the values to
the considered number of individuals.

Title of Lesson 74
ID document: e.g. PPT/L-TCIO-1

Histogram analysis – Example : packages

Go back again on our example : packages.

● Calculation of the arithmetic mean :

● Calculation of the weighted arithmetic average :

It is calculated in the same way as before but used in the case where the
numbers are grouped by class and the values considered at the value of the
class center.

Calculation is for you

Title of Lesson 75
ID document: e.g. PPT/L-TCIO-1

Histogram analysis – Example : packages

● Mode :
This is the center of the class that has the highest absolute frequency or the
highest relative frequency in the histogram.

This is 55kg which belongs to class 5.

● Median :
This is the value that divides the sample size into two equal parts. If the sample
size is even, there are two values in the middle, and the median is the half-
sum of the two values.

Calculation is for you

Title of Lesson 76
ID document: e.g. PPT/L-TCIO-1

Histogram analysis – Example : packages

● Variance :
It is the arithmetic mean of the squares of the deviations of the values of these
observations from their arithmetic mean.

With as parameters :
- ni : class size
- xi : classe center value
- X : Arithmetic mean
- N : Total

The variance is an index that characterizes the distribution (in area) of the
population, i.e. the form of spread of this population in relation to the
average.
Title of Lesson 77
ID document: e.g. PPT/L-TCIO-1

Histogram analysis – Example : packages

To make the calculation easier, we can
construct the attached table. For each class,
the calculation of :

- ni : class size

- (xi – X) : the difference between the class

center and the arithmetic mean

- (xi – X)² : the square of this difference

- (xi – X)²ni : the weighting of the square of

Reminder : X = 56,09 i.e
the difference on the class size
56
Title of Lesson 78
ID document: e.g. PPT/L-TCIO-1

Histogram analysis – Example : packages

We deduce after calculation the value of the variance :

● Calculation of the standard deviation :

This is the square root of the variance.

We deduce after calculation :

Title of Lesson 79
ID document: e.g. PPT/L-TCIO-1

4 - Use and comparison of

these three centering
indicators

Title of Lesson 80
ID document: e.g. PPT/L-TCIO-1

4 - Use and comparison of these

three centering indicators
- Use

- Advantages
A - Mean
- Disadvantages

Title of Lesson 81
ID document: e.g. PPT/L-TCIO-1

4 - Use and comparison of these three centering

indicators – The mean
● Usage:

This is the index used in almost all cases.

• Advantages: • Disadvantages :

- Simple calculations - Very sensitive to abnormal extreme

- Leaves no room for interpretation values
- Concrete meaning - Loss of significance in the case of
- Simple to understand bimodal distribution

Title of Lesson 82
ID document: e.g. PPT/L-TCIO-1

4 - Use and comparison of these

three centering indicators
- Use

- Advantages
B - Mode
- Disadvantages

Title of Lesson 83
ID document: e.g. PPT/L-TCIO-1

Use and comparison of these three centering

indicators – Mode
● Usage:
Elaboration of forecasted times.

• Advantages: • Disadvantages :

- Leaves no room for interpretation in In the case of groupings in classes, the

the case of known unit values mode depends on the chosen
- Simple to define division

Title of Lesson 84
ID document: e.g. PPT/L-TCIO-1

4 - Use and comparison of these

three centering indicators
- Use

- Advantages
C - Median
- Disadvantages

Title of Lesson 85
ID document: e.g. PPT/L-TCIO-1

Use and comparison of these three centering

indicators – Median
● Use:
Balancing (équilibrage) of stations in a production line

• Advantages: • Disadvantages:

- Concrete and easy to understand - Does not take individual values into
meaning account at all
- Simple to define

Title of Lesson 86
ID document: e.g. PPT/L-TCIO-1

4 - Use and comparison of these

three centering indicators

D – Normal law

Title of Lesson 87
ID document: e.g. PPT/L-TCIO-1

Use and comparison of these three centering

indicators – Normal law
In a very large number of cases and under certain conditions that are extremely
frequent in the industrial context, we can consider with little risk of error that
the distribution of a quantity tends to be normal.

This normal law is called GAUSS law or LAPLACE - GAUSS law

The distribution is symmetrical with respect to the central value from where :

Mean = Mode = Median

Title of Lesson 88
ID document: e.g. PPT/L-TCIO-1

Histogram can be surprising… if you don’t know what’s behind

Situation you can meet in industry :

Measurement of a sample of pieces.

Histogram seems not looking like the normal law

The sample of pieces comes from two distinct process:

For example a mold with 2 cavities for a plastic molding

machine.

Pieces from cavity n°1 should not be mixed with pieces

from cavity n°2.

Title of Lesson 89
ID document: e.g. PPT/L-TCIO-1

THANKS FOR YOUR ATTENTION

Title of Lesson 90
ID document: e.g. PPT/L-TCIO-1

Documents Management
Qualité 1 – Chapitre 2 – Rappels et introduction à la maitrise
Title: statistique des procédés ID No.:

Autor: Sébastien Lasserre Date: 30/08/2021

Version: 1.0 Format: ppt

Root:

Description:

VERSION DATE AUTOR DESCRIPTION OF MODIFICATIONS

Sébastien
1.0 30/08/2021 Lasserre Initial document

2,0 29/01/23 Laurent Sestier Added several examples and Galton board experience

Title of Lesson 91