Someone suggest me to have at least two to five different sets (with different bids) for Willingness to Pay (CVM) questionnaire to minimize starting point bias. Is it compulsory to do so? My questionnaire is double-bonded dichotomous.
It is quite standard in CVM surveys to include several different bids for WTP. This has to do more with identifying WTP than starting point bias. In fact, you can always test for starting point bias after the survey has been conducted.
According to Bateman et al. (2002). Economic valuation with stated preference techniques: A manual, p. 384: Strategies to counteract starting point bias: (1) "to tell respondents in advance about the full range of possible values they might encounter" and 2) "to explicitly randomise the first and subsequent values presented to respondents".
