Willingness to pay (WTP) for insecticide-treated mosquito nets was elicited using the bidding game (BG);: the binary-with follow-up (BWFU) and a novel structured haggling technique( SH). Construct validity across question formats depended on the econometric model used. Respondents from higher socio-economic households and those that had suffered a recent malaria attack were more willing to pay. The Tobit model was mis-specified in the BG and SH and the Heckman model was mis-specified in the SH in determining the construct validity of full sample WTP. Ordinary-least-squares was better than the truncated regression in determining the construct validity of positive WTP.
Medical Decision Making (2008) 28 (1) 90-101 [DOI: 10.1177/0272989X07308748]