Minimum Tolerance Level
The function finds a Minimum Tolerance Level (error rate) allowing for a conclusion that two empirical distributions (distributions of WTP or measures of WTP estimated with uncertainty) are 'equivalent' at the 5% significance level. This could be particularly useful for Benefit Transfer (Czajkowski and Ščasný, 2010; Ahtiainen, Artell, Czajkowski and Meyerhoff, forthcoming).
- Equivalence is defined in accordance with Kristofersson and Navrud (2005)
- The optimization uses Two One Sided Z tests (TOSZ, for normally distributed random variables) or Two One Sided Convolutions test (TOSC, for any other distributions)
- The convolutions approach is described by, e.g., Poe, Giraud and Loomis (2005)