Introduction

Smileys and other emoticons as graphic representations of facial expressions are prevalent in everyday life. These visualizations are used frequently in private text messages, chats, and even business emails (e.g., Derks, Bos, & von Grumbkow, 2008; Krohn, 2004; Walther & D'Addario, 2001). Paralleling their increasing use in computer-mediated communication in everyday life, smileys and other emoticons also have been used more frequently in surveys. Recent studies have shown that visual design features and the use of numbers, symbols, and graphics can alter response behavior in web surveys (Christian, Dillman, & Smyth, 2007; Toepoel & Dillman, 2010; Tourangeau, Couper, & Conrad, 2004, 2007). Smileys and other emoticons are likely to convey additional information and, thus, support respondents in answering survey questions. Consequentially, it seems reasonable to use smileys as response scale labels as a means to reduce respondents' response burden and, ultimately, increase response quality. In contrast to this positive perspective, some have argued that replacing verbal rating labels with smileys or other emoticons might confuse respondents or lead to divergent interpretations of the rating scale option meaning, for instance, due to cultural differences in how smiley faces are interpreted (Yuki, Maddux, & Masuda, 2007). The consequence could be an increased response burden and lowered response quality. In light of the increasing use of visual elements in everyday computer- mediated communication, the scarcity of methodological research on the use of smiley face scales in surveys is striking, especially with respect to how such scales may affect the survey response process.