Hi Knicole I generally prefer using respondent totals, but it probably depends on context. You have to ask yourself whether you are interested in the people who took the survey, or the answers to the question. I am generally interested in the people. For example, if I ask the respondents how they commute to work, I am interested in the % of people who bike at all, even if they also use other modes (% of respondents). A different example might be if you asked respondents which of the following lessons they completed from a list of lessons. If you look at the % of respondents, you'd know the % of people who completed a particular lesson. But if you look at the % of answers, you'd know what % of all the lessons completed were in any given lesson. Hope that helps!