In this paper, we show that in Web questionnaires verbal and visual languages can be used to create groups and subgroups of information, which influence how respondents process Web questionnaires. Following Schwarz (1996; and also Schwarz, Grayson, & Knäuper, 1998) we argue that respondents act as cooperative communicators who use formal features of the questionnaire to help guide them through the survey conversation. Using data from three Web surveys of random samples of Washington State University undergraduates, we found that when response options are placed in close graphical proximity to each other and separated from other options, respondents perceive visual subgroups of the categories, increasing the likelihood that they select an answer from each subgroup. We also found that graphical proximity creates subgroups with and without the use of category heading to describe the subgroups and that the addition of a verbal instruction to “please select the best answer” encouraged respondents to select one answer from each subgroup instead of overriding the effects of proximity. In addition, the effects of grouping were consistent when the subgroups were positioned either vertically or horizontally in relation to each other. Lastly, we found that the effects of visual grouping are consistent across both opinion- and behavior/fact-based questions, although the effects appear to be greater on opinion-based questions. Our findings contribute to the increasing evidence that both verbal and visual languages influence how respondents process and respond to surveys. Because respondents interpret the verbal and graphical features of survey questionnaires as elevant to answering the survey, inadvertent or stylistic design changes can influence how respondents process and respond to survey questions.