g GE Global Research

Green Belt DMAIC Workshop

BASIC STATISTICS
Part 1

Define
Measure
Analyze
Improve
Control

i 1
GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.1
 g GE Global Research

The 12-Step Process

Step Description Focus Tools SSQC Deliverables

Define
A Identify Project CTQs Project CTQs (1)
B Develop Team Charter Approved Charter (2)
C Define Process Map High Level Process Map
(3)
Measure
1 Select CTQ Y Fishbone, FMEA, Pareto Project Y (4)
Characteristic Customer, QFD
2 Define Performance Y, X Customer, Blueprints Performance Standard
Standards for Project Y (5)
3 Establish Data Y, X Gage Study Data Collection Plan &
Collection Plan, MSA (6), Data for Project
Validate Measurement Y (7)
System*, & Collect Data
Analyze
4 Establish Process Y Capability Indices Process Capability
Capability for Project Y (8)
5 Define Performance Y Team, Benchmarking Improvement Goal
Objective for Project Y (9)
6 Identify Variation Sources X Hypothesis Tests Prioritized List of all Xs
(10)
Improve
7 Screen Potential Causes X DOE-Fractional, Process List of Vital Few Xs (11)
Map, Fishbone, FMEA
8 Discover Variable X DOE-Full, Prediction Proposed Solution (13)
Relationships Eqns., DFSS,
& Propose Solution Statistical Tolerancing
9 Establish Operating Y, X Piloted Solution (14)
Tolerances & Pilot Solution
Control
10 Validate Measurement* Y, X Gage Study MSA*
System
11 Determine Process Y, X Capability Indices Process Capability Y, X
Capability (15)
12 Implement Process X Risk Analysis, Sustained Solution (15),
Control System Mistake Proof, SPC Documentation (16),
& Project Closure QPT Leveraged Solution (17),
* You may want to validate the Measurement System throughout the DMAIC steps. Financial Audit (12)

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.2

 g GE Global Research

Learning Objectives

 The Nature of Statistics

 Types of data
 Mean, Variance, Standard Deviation
 Scatter Plot
 Box Plot

3
GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.3
 g GE Global Research

What is statistics?

• The data

• Science of organising, catagorising and analysing

the data.

What kind of data do you deal with?

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.4

 g GE Global Research

Types of statistics
Descriptive Statistics :
Numerical index that describes some
characteristics of the population.

Market share of Competitors

1
2
3
4
5
6

Inferential / Analytical Statistics :

Quantitative technique that enables the experimenter
to generalise about the population using limited
number of observations.

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.5

 g GE Global Research

Two basic types of Data

Continuous data
Data that characterizes a product or or process feature in
terms of its size, weight, volts is said to be continuous by
nature. In other words, the measurement scale can be
meaningfully divided into finer and finer increments of
precision.

To apply the normal distribution, one must necessarily use

continuous data.

Discrete (attribute) data

Another way to look at the data is to merely count the
frequency of occurrence—for example, the number of
times something happens or fails to happen. Notice that
such data is not capable of being meaningfully subdivided
into more precise increments, and therefore is said to be
discrete by nature.

The Poisson and binomial distributions are used in

connection with this type of data.

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.6

 g GE Global Research

Two basic types of Data

Data Type

Variable Data
Attribute (discrete)
measurement
count or classification
on continuous scale

classify
count
as defective
defects
or not

normal or other
binomial Poisson
continuous distribution
distribution distribution
(e.g., Weibull for life data)

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.7

 g GE Global Research

Continuous Data - Statistical Terminology

Population – a complete set of data. All items of interest.

 Number of elements in a population is denoted by N

Sample - a subset of data from the population

 Number of elements in the sample is denoted by n

Mean - the average of the population or sample

 Mean of a population is denoted by 
^
 Mean of a sample is denoted by X or 

Variance - the average squared difference from the mean. (It represents the
spread of the data.)
 Variance of a population is denoted by 
 ^
Variance of a sample is denoted by s2 or 

Standard Deviation is the (positive) square root of the variance. (It also
represents the spread of the data.)
 Standard Deviation of a population is denoted by 
 Standard Deviation of a sample is denoted by s or ^ 

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.8

 g GE Global Research

Statistical Terms and Definitions

Mean - The average of the population or sample.

Denoted by x or ^
 for a sample and  for a population.

n
Formula for the Mean
 xi
1
x = , where x1  first point and
n
xn is the last point of a sample.

Example: Given a sample: {1,3,5,4,7 }, the mean or

average is;

x = (1+3+5+4+7) = 20 = 4.0
5 5

The
TheMean,
Mean,or
orAverage
Average((xx))
of
ofthe
thesample
sampleis
isequal
equalto
to4.
4.

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.9

 g GE Global Research

Statistical Terms and Definitions

Standard Deviation - A measure of the spread of the

data. Usually denoted by  for a population and s or ^ for
a sample.

N
2
 ( Xi -  ) Formula
Formula
=
i=1
for
foraaPopulation
Population
N

n
2
 ( Xi - X ) Formula
Formula
S=  = i =1
for
foraaSample
Sample
n-1

Variance - Average squared difference from the mean.

Usually denoted by s2 or ^2

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.10

 g GE Global Research

Calculation Exercise

Calculate the sample mean, standard deviation,

and variance associated with this data set.
98, 98, 90, 104, 106, 110, 101, 104, 102, 96

x x x ( x  x )2
1
2
3
4
5
6
7
8
9
10

X
 2

n n
 Xi  ( X 1  X )2
x i 1
n
  i 1
n 1

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.11

 g GE Global Research

Problem Solving with Statistics

Typical difficulties faced by an investigator

• Experimental error (systematic and random)
• Confusion of correlation with causation
• Complexity of the effects studied

How to use statistics - Think before doing!

• Find out as much as you can about the
problem.
• Don’t forget non-statistical knowledge and
understanding.
• Define objectives.
• Define how you will measure that you have
achieved the objectives
• Plot your Data! Look at it and think!

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.12

 g GE Global Research

Data Depiction

Graphical tools are very useful for taking a

“first look at our data sets”. We may quickly
see trends or problems that will help us
organize our approach to a more complete
analysis of our problem.

Useful Graphical Tools

Scatter Plot
Box Plot
Histogram
Frequency Polygon

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.13

 g GE Global Research

Scattergram / Scatter Plot

Purpose: To study the possible relationship between one variable
and another

How: Plot the data

Analyze data to determine correlation (straight line or
tight clusters indicate strong relationship)

Cause and Effect Plot

2.5

1.5 Effect 1
Unknown

0.5

0 0.5 1 1.5 2 2.5 3 3.5

Independent variable

The scatter plot helps us to visualize the relationship

between two variables

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.14

 g GE Global Research

Graphical Tools

Box Plot
 Enables quick visual comparison of data
sets

 Useful for report outs

 Useful for data segmentation and

stratification

 Remember to back up any conclusions

with appropriate hypothesis tests (See
Basic Statistics 3)

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.15

 g GE Global Research

Box Plot Concept

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.16

 g GE Global Research

Box Plot in MiniTab

Is the average distance resulted from Operator 1

different from that from Operator 2? A hypothesis test
would be the ideal tool to answer the question.
However, can you show me the difference graphically
for a quick view?

MINITAB FILE: Catapult.mtw

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.17

 g GE Global Research

Box Plot in MiniTab

MINITAB FILE: Catapult.mtw

Use Box plot to compare the distribution between Operator 1 and 2.

GB DMAIC - Basic Statistics Part 1 Version 2.0 5/2002 7.18