I'm conducting a market segmentation involving clustering respondents
based on their responses to attitudinal statements contained in an
attribute battery where all attitudinal statements are answered on the
same 5 point scale (strongly disagree to strongly agree). Respondents are
being grouped into segments using latent class analysis. I'm looking for
information regarding the following:
(1) Does SPSS have a module to conduct latent class analysis?
(2) A point of view/recommendation regarding using individual
intercorrelated attitudinal attributes versus orthogonal principle
components for the same set of intercorrelated attitudinal attributes as
input basis variables into latent class cluster analysis. Which strategy
is better (or are both not advisable)?
(3) Strategies for dealing with response scale usage bias (high rater
segments and low rater segments due to scale usage bias) within latent
clas analysis. Can latent class cluster analysis be run on mean-centered
attitudinal ratings? Is there a better strategy for removing the effects
of scale usage bias (and if so, are there references that can be suggested
explaining how to run these analysis)?
Thank you.
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
