Since some years I am working with discrete choice experiments and I am often doing analysis with Sawtooth Lighthouse Studio. However, when it comes to further steps of analysis, at some point you need to think about choice simulation if you want to get insights into choice scenario details (shares of preference). I recently have heard (can't remember where) that every kind of choice simulation is not reasonable as it stretches people choices to its edges and the result may be more a randomized guess than trustworthy prognosis. On the other side there is a lively discussion within the Sawtooth forum and the developers (Brian Orme and others) claim that it is anyhow very reasonable. As I did not find any discussions or paper beside Sawtooth's forum I wanted ask what people here think about it. Having said that, I am not intending to start a "technical" discussion but more a practical one.
Greetings, Boris
- Badges
- Science topic
- Similar topics
- Correction