[R] Please help: Latent class analysis with poLCA

Janina_84 Janina.Hofmann at kuei.de
Thu Oct 9 17:26:47 CEST 2008

Dear R-experts,

I am currently writing my theses about what - else but price - drives low
cost carriers` buying behavior.

To evaluate the collected choice-based conjoint data, I have to perform a
latent class analysis.
I have found the package poLCA, but somehow I don`t get the output I need...

According to what I have read about latent class analysis so far, the
LC output should contain

1) The relative size of the segments (which I obtained with poLCA)
2) The segementspecific parameters for each latent class/segment (in order
to see if and how they differ in their drivers of buying behavior)
3) A classifiaction of participants as to which class the belong with a
certain probability (to be able to assign them to the clsses and profile
the segment)

Unfortunately, I cannot find the 2) and 3) information in the poLCA-output,
and I don`t know how to deal with the conditional response probabilities I
obtain with the package- can they maybe be used to obtain
the segmentspecific parameters? Or is the program generally not designed to
do this?

Another problem is that I don`t know in which form the input file needs
to be, since none of the examples deals with choice-based conjoint data. I
have data in the following form, assumed 4 attributes, each with
two levels,  and each choice can be described by the chosen alternatives

ID A1 A2 A3 A4
1  1  2  2  1
1  1  1  1  2
1  2  2  1  1

Do I maybe need to perform frequency counts?

I am a real "greenhorn" concerning R , so I am
quite desperate and hope for your help.

Thank you so much!


View this message in context: http://www.nabble.com/Please-help%3A-Latent-class-analysis-with-poLCA-tp19901303p19901303.html
Sent from the R help mailing list archive at Nabble.com.

More information about the R-help mailing list