New Frontiers

Abstract:A Bayesian network’s ability to graphically model cause and effect relationships between variables can be applied to customer satisfaction surveys to show how the distribution of responses to one question affects the responses to another question. The model can be used for diagnostic and prediction purposes and to provide management insights. Two examples illustrate how the Bayesian network is applied to data collected from annual customer satisfaction …

Access this article
Other ways to access this article

Social Bookmarking

Digg, delicious, NewsVine, Furl, Google, StumbleUpon, BlogMarks, Facebook

It is a very nice method to connect the main attributes and make the right decisions for the business based on survey-structured facts.

--Fabricio Lopes, 08-18-2009

These Bayesian principles are at least 25 years old, used by StrategicVision and the Wirthlin Group to manage voting behavior and product and service satisfaction.

The basic problem is using a satisfaction scale. Those scales are "tired" - used up with little functional discrimination among the data points near the top of the scale.

Using a score of 4 or 5 kills the discrimination even more. The linear networks mimic our more comprehensive ValueCentered work from attributes to benefits to values/emotions to images - related to behavioral outcomes like advocacy, purchase intent, or even actrual re-purchase.

The complex maps mirror the Baseyian maps we developed with Wirthlin in 1988-1989 through 1992.

If you want to know my credentials either check or

Dr. Darrel Edwards, Ph.D.
ABPP - Diplomate American Board of Professional
Senior Member-ASQ
Fellow AACP - American Academy of Clinical
Chosen as One of the Top 2000 Scientists of
the 20th Century - Cambridge
International BiographicaL Centre

--Dr. Darrel Edwards, 08-10-2009

Featured advertisers