Simple question on hypothesis test

["Data Analytics Corp." <dataanalytics@earthlink.net>]

Hi,

I have a simple, naive question.  A client did a taste test of two 
products, say A and B.  Consumers were asked to choose one.  The 
proportion choosing A is PA = 0.45, and so PB = 0.55.  The client wants 
to know if PA is significantly different from PB.  How is this tested, 
and does it even make sense?  A simple t-test doesn't seem appropriate.  
My first reaction was that I could test, say, PA against a standard of 
0.5 to see if there is a random selection or if people really did not 
prefer A.  If they really did not prefer A, then PA does differ from 
PB.  Any comments?
Sorry for the dumb question, but this is what clients ask.

Walt Paczkowski

--
________________________

Walter R. Paczkowski, Ph.D.
Data Analytics Corp.
44 Hamilton Lane
Plainsboro, NJ 08536
________________________
(V) 609-936-8999
(F) 609-936-3733
dataanalytics@earthlink.net
www.dataanalyticscorp.com