Date: Mon, 2 Dec 2002 12:21:52 -0800
Reply-To: Becky Wu <firstname.lastname@example.org>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Becky Wu <email@example.com>
Subject: SPSS Conjoint Market Share Simulation
Content-Type: text/plain; charset="iso-8859-1"
I am simulating market shares for three potential products using SPSS. The simulation methods I used are Maximum Utility, BTL and Logit. Interestingly, market shares of the three products generated by BTL and Logit are in the same order in terms of size. That is, product #3 has the largest market share, and product #2 has the second largest market share, and product #1 has the smallest in both methods. However, the result from Maximum Utility is somewhat different - product #3 is still number one, but product #1 is now the second largest and product #2 the last.
I understand the algorithm for these three methods are different, and therefore I didn't expect exact numeric values for the market shares generated by different methods. But I thought the order of the market share sizes should be the same. I would greatly appreciate any insight on this seemingly conflicting result.