[Home Page] [Previous Page]
[Old Roundtable Archives] [New Roundtable]
Posted by Ian Straus on January 27, 2000 at 11:13:50:
In Reply to: Conjoint - free software, algorithms? posted by Mike Pritchard on January 26, 2000 at 16:36:58:
You said "no survey, just getting one or two opinions". I take it you mean to gather the opinions of one or two PEOPLE, because there's no way you can do conjoint without a survey, it's a contradiction in terms. But you want to do it really small and cheap....time is no object?
Well, if first you design your survey - essentially design the set of choices you intend to present to people,
1) You want a balanced orthogonal array of choices. Designing this is one of the benefits of the conjoint programs. But you could get the design from a table in a book. That's what people did before the software was so common.
2) You could then ask your questions, get your rankings, and evaluate them on an individual basis by doing one regression run per person.
I haven't tried this with the regression features or add-ins of Excel; but there's no obvious reason it shouldn't work.
I don't think you could do it by hand.
You might get someone with one of the conjoint programs to help you with the design.
Subject: