One of the recommendations given by the NOAA panel to CVM practitioners is allowing survey respondents to choose 'no-answer' or 'don't know' option. The improved CV methods allowing those options commonly have used revised ordered probit models where the constant term of the WTP function is identified. Recently, King et al.(2004) have suggested a way of reducing the ambiguity in respondents' answers in research survey providing anchoring vignettes to them. Our study combines the multiple response options CV study with the anchoring vignette models. The identification issue of the suggested anchoring vignette CV model is addressed. The anchoring vignette model is applied to an issue of paying for public R&D investments.