Discrete Choice Modeling

作 者:Warren F. Kuhfeld  阅读次数:7462次  发布日期:2005-07-08


This report shows you how to use the multinomial logit model (Manski and McFadden, 1981; Louviere and Woodworth, 1983) to investigate consumer’s stated choices. The multinomial logit model is an alternative tofull-profile conjoint analysis and is extremely popular in marketing research (Louviere, 1991; Carson et. al.,1994). We will discuss designing a choice experiment, preparing the questionnaire, inputting and processing the data, performing the analysis, and interpreting the results. Discrete choice, using the multinomial logit model,is sometimes referred to as “choice-based conjoint.” However, discrete choice uses a different model from fullprofile conjoint analysis. Discrete choice applies a nonlinear model to aggregate choice data, whereas full-profile conjoint analysis applies a linear model to individual-level rating or ranking data. This report is TS-677E, the August 31, 2002 edition for SAS Version 9.0. It is a revision of the April 1, 2001 report for Version 8.2 and other previous editions. This edition uses SAS macros and features that are new in Version 9, including the new experimental design macro, %MktEx . This report heavily relies on a number of macros and procedures.

[浏览文章详细内容]


关闭】 【打印