I have prepared a CBC design with below attributes:
Brand : 13 levels
Price : 17 levels
There is brand A with 3 price levels (1-3) ,Brands B,C,D,E with 6 price levels (4-9) and 8 brands F,G,H,I,J,K,L,M each of these brand have own fixed price. Respectively price level 10,11,12,13,14,15,16,17, where brand F appears only with price 10, brand G with price 11 etc.
It required some manual work with the design file to avoid prohibitions but my question regards analysis.
After collecting the data Im going to recode attribute Price onto 3 price attributes:
Price1 for Brand A
Price2 for Brands B,C,D,E
Price3 for rest of brands with fixed prices.
and estimate utilities by HB.
I think that utilities for the attribute Price3 can be ignored for shares of preference as whole information about products F - M are already captured by brands utililies since each brand was always shown with just 1 price. So when recoding .cho file I could assign to Price3 just zero.
Is it justified to ignore utilities for Price3?