VOLUME 1, ISSUE 1
DUAL RESPONSE™ CHOICE MODELS
Click here for PDF download.

When you do conjoint studies, do you include “none” or “no-choice” in the questions? This is a smart thing to do - it is very natural for the respondent and it is the only way you can make predictions about how many new people will purchase in your category when you introduce a new product.  
 
But did you realize that if you do this, you could be wasting valuable respondent time on questions that will tell you next to nothing about how respondents choose products? If you think about it for awhile - which is what we like to do - when someone answers “none”, you learn almost nothing about what what would motivate them to buy. In the worst-case scenario, the respondents chose “none” all the time and you just wasted money collecting data that tells you nothing about what you can to do to make your product better.   
 
So, what is the solution?  Well, in 2003 we introduced Dual Response choice questions to the market research community.  The idea is clever and powerful.  Instead of letting respondents get away with only telling you “no, I don’t like any of these options,” ask them two questions.  “Which would you buy if you absolutely had to?” and then “Would you really buy it?”. We showed, in a paper published in Marketing Letters, that using Dual Response resulted in more accurate models - the equivalent of increasing your sample size by 20-30%.  
 
So, when should you use Dual Response tasks?  Well, since it takes almost no additional respondent time, we use it on almost every choice study we do.  But, it is particularly important when you have difficulty recruiting respondents and don’t want to waste precious time with them or when your study involves a new product offering  where people may choose “none” frequently.