Does gender-specific starting point bias in choice experiments prevail among well-informed respondents