Conjoint Analysis, Syntax, Cards

[GentlemanEddie]

Dear all,

I'm setting up a customer research using a conjoint analysis (SPSS) and have a few questions.

1. Do I need to set up the syntax file before collecting answers or can I do that afterwards?

2. Is there a way to change the orthogonal design (in SPSS) after I've run it? Or do I have to punch in the values all over again everytime I want to add/remove a level or factor?

3. Currently, when running the orthogonal design I get 16 cards (3-4 factors, 2-5 levels). Unfortuately, this is too many for the subjects to rate (I'm thinking of running a ratings-based conjoint). I realize that I will loose statistical power by reducing the cards,
a) but if I clear some cards from the data view, will that "screw up" my running the analysis?
b) Is there a way to force SPSS to only create an X number of cards maximum?

Thanks very much in advance!

 

[Injektilo]

It's unfortunate you have to use SPSS, because you would save yourself a ton of hassle if you could use Sawtooth software's SSI Web instead.

Having said that, 16 ratings is not all that bad. The rule of thumb is no more than 30.

I'm not able to answer your other questions, as I'm not used to running conjoint analysis in SPSS, but you will likely to re-launch your design if you drop a level or an attribute.

 

[Matilda]

Hi,
i have the same problem. SPSS created 16 Productcards for my analysis. Unfortunately this is not manageable in my survey. Therefore I would like to know whether there's a possibility in SPSS to limit the numer of product cards to a certain number (just as it's possible to determine a minimum number of cards).

As I have not found any answer to this, there might be no solution in SPSS. Therefore I have created an reduced design myself with 9 stimuli, which is much better for my survey. Can I import my own reduced design in SPSS? How is it done? And can I create holdouts card without using SPSS Orthoplan?

I would really appreciate hints to solve my problem! Thank you!!
Matilda