The bias of the questions should be apparent. By asking "is it ok to punch Nazis", it never actually ask the respondent what constitutes a Nazi. Are we talking

The bias of the questions should be apparent. By asking "is it ok to punch Nazis", it never actually ask the respondent what constitutes a Nazi. Are we talking about people who are actively murdering minorities? How do we know what the respondent thought the survey meant when they use words like that?