choice-metrics.com

Dear all,

Thank you for this forum!

I am new in choice experiments and ngene. I am trying to design a labeled choice experiment with three different alternatives and one no buy option, 3 attributes and 3 levels for each attribute. Price is one of the attributes and the only one that should have 3 different levels for each alternative.

One of the problems is the reference price which differs between people. One solution I would think of is the heterogeneous pivot design. However, I am trying to understand if there is a possibility to do this design for the labeled choice experiment. In the manual I see example of unlabeled choice.

Thank you very much in advance for your response!!

I generally generate a separate design for each population segment (e.g., with a different reference price) by simply running Ngene multiple times with different levels. Using absolute or relative pivots is generally tricky to implement in a survey instrument, so the easiest is the generate multiple designs and pick a design based on information provided by the respondent. Does that answer your question?

Michiel