According to the Merriam-Webster
dictionary, synergy means a mutually advantageous conjunction
or compatibility of distinct business participants or elements
(such as resources or efforts). In layman’s terms, it
means that the sum of the parts is greater than the whole. As
any chemist could tell you, synergy is a valuable thing to
have. JMP and chemistry have this synergy – here are a few
that come to mind.
For any chemistry-related project,
there are unique features in JMP to help you analyze your
data. Start with analytical chemistry. How do you know that
you are getting reproducible results from your measurement
device? JMP has advanced tools for evaluating your measurement
process, allowing you to pinpoint and quantify the sources of
variability in your measurement system.
processes need to be monitored to ensure that they are
producing quality products that are in specification. JMP has
the tools for monitoring these processes. With the Control
Chart Builder, analysts can drag and drop the variables they
want on the chart and JMP will determine the appropriate
control chart. Find an out-of-control condition? JMP’s dynamic
linking allows you to quickly find all of the information
related to it and then find a remedy.
understand complex chemical reactions or to optimize the way a
plant is operating, design of experiments is a critical tool.
Designing experiments is one of JMP’s strengths. The Custom
Designer will quickly allow a researcher to plan experiments
for complicated scenarios and then help plan the analysis by
building an appropriate model for analyzing the experiment.
Plus, JMP provides all of the tools necessary for
understanding the experimental results and optimizing
There are so many other ways that chemists
and chemical engineers can benefit from using JMP software. So
many, in fact, JMP has a website dedicated to chemists.
And if you are interested in design and
analysis of mixture experiments, want to learn how to analyze
multidimensional data AND are attending the JMP Discovery
Summit in September – sign up for one of our pre-conference
I’ll leave you with this
Q: What did the bartender say
when oxygen, hydrogen, sulfur, sodium and phosphorous walked
into his bar?
A: OH SNaP!
P.S. If you haven’t signed up already and would
like to continue receiving JMP Training: News and
Views, please sign
up to be on our interest list.
In my 30 years of applying
design of experiments methodology, how to efficiently
investigate a process or system in a reasonable number of
experiments, given the time, material or cost constraints was
a frequent challenge. Many times, we would have to prune our
brainstormed factors down to six or seven to stay within our
budget for the number of runs that we could afford.
Unfortunately, that prevented us from considering some
potentially important factors and exploring how those factors
were affected by the other investigated factors. Thanks to a
brand-new class of screening designs by Bradley Jones and
Christopher Nachtsheim called definitive
screening designs, this is no longer the case.
definitive screening design requires only 2N+1 experiments.
For example, a 10-factor investigation requires only 21
experiments. In addition, these experiments can delineate main
effects clear of confounding with two-way interactions and
often can even delineate polynomial terms. In contrast, a
classical design of experiments would often include
center-point replication, but this would give only a global
assessment of the lack of fit. On the other hand, the
construction of the definitive screening design can often
identify the active polynomial term.
A paper by Jones
and Nachtsheim on definitive screening designs won the
prestigious ASQ Brumbaugh Award because it made the largest
single contribution to the development of industrial
application of quality control in 2011. In addition, the first
application of this new design type by Novomer in 2012 was
recognized by the ASA with the Statistics
in Chemistry Award.
Below is a color map of
the correlations for a 10-factor definitive screening design.
All of the main effects are clear of confounding, with all
two-way interactions and polynomial terms.
all, this new class of designs is available to you on the JMP
File Exchange as an add-in.
This tip was
provided by Louis Valente, JMP Senior Manager, Technical
up to continue receiving JMP Training:
News and Views.