Within web survey design, the use of HTML (check-all-that-apply) boxes as opposed to radio buttons allows respondents to select multiple answers to one question. For such multiple response questions, research suggests that respondents select the response options differently (e.g., Christian and Dillman 2004). For example, while vertically listed response options seem likely to be selected from top to bottom, multiple column listing of response options coupled with the landscape orientation of computer screens may encourage respondents to select horizontally. Inspecting the sequence of response selection may shed valuable light on the ways in which respondents answer multiple response questions. Knowledge of such behavior may be more important now than ever before. Recent research has shown that nearly 20 percent of mobile phone users do the vast majority of their web browsing on a smartphone as opposed to a computer or tablet (Miller et al. 2012). This percentage is likely to increase as smartphone ownership has now surpassed basic cell phone ownership. Designing multiple response questions for web surveys that require very little space yet do not require scrolling, which can be cumbersome on a phone, makes the use of horizontally arrayed response options attractive. However, if people select options horizontally, instead of vertically, the distribution of answers may be affected. Thus, in this paper, we ask whether the orientation of a multiple response question affects the sequence of response selection and, if so, does this have an effect on the substantive distributions?