Clueless in need of Help in experimental design and analysis method

#1
Hi guys,
I need to plan and submit a small report, in a customer behaviour course that haunts me over a 2nd year already...

I need to evaluate if the numerical representation I am using for some product's spec. will affect the custumer willingnes to buy it (quantitative parameter). The scale is my main concern - narrow scale vs wide scale ( e.g. warranty in years vs warranty in months..etc).

I also need to add and test a qualitative parameter , and it's effect on willingness to buy the product, and at the end - do the two parameters interact ( 1st quantitative with the 2nd qualitative,,,)

How should I plan my survey?
What is the correct way to analyze the data ?

Please give a brother in need a hand....
(I'll be Happy to add more details I'd needed - I'm desperate ;( )

Thanks
 
#2
I was thinking on showing the two products with their stats and ask to rate each of them, on a 1-10 scale according to the willingness to purchase each. 1 group will see the narrow scale stats (1% active ingredient vs 3% active ingredient) and 2nd group will see wider scale (1000 active units vs 3000 active units). I was thinking to use a t-test on the avreges inside each group. What should I use to compare the two groups? ( If we'll see adifference in the reaction to the scales)?
 

hlsmith

Not a robit
#5
Please write out your question so we know what variables to think about. If you need a continuous, categorical, and multiplicative interaction for whether a purchase is made - this sounds like multiple logistic regression. We don't know your content area, so we can't tell you what variables to use. Of note, if you collect warranty days as a predictor, this can easy be converted to a more macro-level variable later on. This is commonly done in logistic regression, you can enter it as days or month and look to see which is more meaningful, likely the latter.