Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.


In this paper (A discrete choice experiment to explore patients’ willingness
to risk disease relapse from treatment withdrawal in psoriatic arthritis, Clin Rheumatol (2016) 35:2967–2974), Authors computed the willingness-to-risk in the following way. "Marginal rate of substitution or the tradeoff between attributes and levels was estimated by dividing one coefficient by another. This was used to estimate the maximum acceptable risk (MAR) in relapse that respondents were willing to accept (i.e. the mean level of risk in negative outcomes) for a given improvement in benefit outcomes. It was calculated as the change in risk of relapse that would exactly offset the perceived benefit of an improvement in sickness/nausea or health status. Therefore, the MAR represents the level of non-inferiority in the risk of relapse from the patients’ perspective." Looking at theri calculation, it seems too simple and intuitevely I have some doubts about their procedures. What is your opinion?
asked May 16, 2020 by Patopas (155 points)

1 Answer

+1 vote
Best answer
This is a pretty standard of estimating WTR in the literature.  Just like willingness to pay (WTP), however, we'd recommend doing this not with the utilities themselves, but in a simulator, within a realistic competitive context.  If one really wanted to do this algebraically, one would probably be better off using the more complex method suggested by Allenby and several coauthors:    

Allenby, Greg M,Jeff D Brazell, John R Howell, Peter E Rossi. 2014b. Economic
valuation of product features. Quantitative Marketing and Economics 12(4) 421–
answered May 16, 2020 by Keith Chrzan Platinum Sawtooth Software, Inc. (111,275 points)
selected Apr 3 by Patopas