Conjoint Choice Model

[Yannicdonleo]

Dear users,

For my profession i conducted a survey amongst customers of a electronic brand to estimate the willingness to pay (WTP) for 3D television.
Each of the 800 customers did get a survey in which they got a card with the description of 3 different products including its price and they had to choose their most preferred choice.
Al these cards where randomly created using 4 different attributes that specified the product: size screen small medium or large, quality of screen high or low, 3D television yes or no and price 600, 800, 1000, 1200 or 1400.

For example a customer had to choose between the following 3 products:
1 large screen, low quality of screen, 3D television, price 1200
2 small screen, low quality of screen, no 3D television, price 800
3 large screen, high quality of screen, 3D television, price 1400

My question now is: using all the collected data of the 800 surveys (in which for every customer te card was randomly created), how can I estimate the WTP for 3D television.
Can you guys point me in a direction? I already looked at multinomial regression models but my problem is that such a model can only give an estimate of the amount of increment in likelihood of an aspect of an attribute, not the decrease in likelihood for for example 1 dollar extra in price.
For example: a price of 1400 decreases the chance a customer buys a product with approx 5%.
What I want: one dollar decreases the chance that a customer buys a product with approx X%.
If I know the increment of chance that a customer buys a product when it has 3D television, I can calculate the WTP using X.

In advance, thanks a lot for reading and hopfully helping!