sbathe@roper.com wrote: > > Hi there! > > I'm not sure I understand your problem correctly. Are you trying to see > how individuals in groups 1-5 would classify into groups 6-10? In that > case you can perform the discrim analysis on groups 6-10 and then use the > Fisher scores to classify the respondents from groups 1-5. > Hope this helps.

Hi,

I only used groups albels 6-10 to keep trace of the actual groups that they belong to (1-5), and in the same time, to eliminate them from the analysis (telling SPSS that the range of grouping var . is 1-5).

Levente

-- **************************************************************** * Levente Toth (levi@cis.plym.ac.uk) - Centre for Intelligent * * Systems, University of Plymouth, Drake Circus, Plymouth * * Devon,England WEB: www.cis.plym.ac.uk/cis/levi/home.html * * Tel: 44-1752-233514, 233516 / Fax: 44-1752-232583 * ****************************************************************