The impact of visual features of questions on survey responses may be more marked in web surveys than in self-administered paper questionnaires, since web surveys often include richer visual material. Existing research on web surveys shows that respondents in web surveys may use visual cues to help interpret the questions or answer options. For example, respondents may make inferences about the conceptual range covered by a response option based on its height or width. Options that take up more space on the screen may be seen as representing a broader conceptual range than those encompassing less space. As a result, we predicted that respondents would be more likely to select a response option when it appears wider on the screen. We tested this prediction in an experiment embedded in a web survey. The experiment presented the response options for a set of questions either vertically or horizontally; in addition, the response options were spaced evenly, with gaps between two of the options, or with one of the options presented on multiple lines — that is, some response options took up more space on the screen than others. This presentation presents the findings from the results from this experiment as well as the results from five similar web experiments that examined the effects of height, width, and spacing on responses in web surveys. In addition, practical guidance on the design of response scales in Web surveys based on the results from these surveys will be provided.