 |
|
Is your QFD math up to date?
QFD's renewed popularity in the United States is due, in
part, to its increased use in Six Sigma and Design for Six
Sigma. Most readers know that this Six Sigma is rich in
statistical tools to provide the accuracy necessary to
achieve 3ppm levels of quality.
It is time, then, that QFD practitioners address the
issue of the numerical inaccuracy of the QFD matrices.
Historically, QFD began at a time when even four-
function calculators were unknown. Early Japanese
practitioners made their charts manually and often used the
letters a, b, c to determine importance and other measures.
As simple calculators became available, numbers became
easier to manipulate and so were used more and more. Since
customer needs and functional characteristics had different
scales of measurement, it was hard to compare them, and so a
simple 1-5 rating scale was adopted to keep all the data in
a comparable scale.
|
|
Ordinal scale, suspect math |
|
|
|
The problem is that this 1-5 rating scale is an ordinal
scale. The QFD operations performed in the Quality Planning
Table, such as the Customer Importance and Competitive
Assessments are suspect.
Is a rating of 4 twice as important as a rating of 2 for
all the customer needs, or could it be different? With an
ordinal scale we cannot tell. The Improvement Ratio where we
divide the Plan by the Current level is improper math
because you cannot divide ordinal scale numbers.
The Sales Point, too, is an ordinal scale, and it is
equally improper math when we multiply the Customer
Importance x Improvement Ratio x Sales Point to calculate
the Absolute Weight and Customer Needs Weight because
multiplying ordinal scale numbers is also improper math.
Then, we multiply the Customer Needs Weight by the
Relationship Strength (1, 3, 9 is also an ordinal scale),
and then sum and divide again for Functional Characteristic
Weights.
Can anybody really know what these weights mean?
|
|
Analytic Hierachy Process |
|
|
|
To increase the accuracy of QFD numbers for better Six Sigma
compatibility, we should also consider using ratio scale
numbers.
Fortunately, such a method exists and has been used in
QFD since the late 1980s. It is called the
Analytic
Hierarchy Process or AHP for short. It was developed by Dr.
Tom Saaty and is one of the most rigorous tools used in QFD
today.
The AHP has added benefits in that it can capture
priorities using natural language comparisons and convert
them into ratio scale numbers. The process can be done with
a calculator, a spreadsheet, and even with dedicated
software.
|
|
Bring your QFD math to Six Sigma level |
|
|
|
This approach is now an integral part of the QFD Institute's
QFD Green Belt® and
QFD Black Belt® programs.
These programs are tailored to the unique needs of each
company and project. If you want to scout these courses
first, several public courses are available but we use
generic models. Subsequent tailoring is then recommended.
More can be learned by contacting the
QFD Institute, +1
734-995-0847 (Mon-Fri, 9am-5pm, US Eastern Time).
© 2006-2008 QFD Institute.
|
|
|
