A longitudinal study of emoticon use in text messaging from smartphones

Abstract

Our goal in the present study was to understand how emoticons are used in text messaging and, in particular, how genders differed in the frequency and variety of emoticons used via this medium. Previous research has found small and sundry differences in emotive expression online suggesting that technology has closed the gender gap. However, the data collected in these studies were public. In this study, we collected real portions of private communications data from individuals’ smartphones over a 6-month period. SMS messages, in general, were not used very much overall, with only 4% of all messages containing at least one emoticon. Still, differences between genders manifested in the amount and variety of emoticons used. Females sent more messages with emoticons; however, surprisingly, males used a more diverse range of emoticons.

Introduction

Text messaging (Short Messaging Service; SMS) has become an important mode of communication throughout the world and is increasing at a rapid rate, with an estimated eight trillion text messages to be sent this year alone (Global Mobile Statistics, 2011). In general, users employ this medium to coordinate activity, maintain social relationships, fill dead time, and share information with others in their social network (Ling, 2005). For teenagers, text messaging is the most popular way to communicate with their social networks, exceeding face-to-face (F2F) communications, emails, and voice calls (Pew Internet & American Life Project, 2010).

Our interest in the current study is the use of emoticons via this modality. Similar to other types of computer mediated communication (CMC), users can include emoticons within text messages to provide socioemotional context. These visual cues have been noted as the primary way to express emotion in CMC (Riva, 2002) and a way to replace non-verbal communications when not F2F (Walther & D’Addario, 2001). Most studies that have examined emoticon use have used information found on public portals via another medium (e.g., instant messaging, Derks, Bos, & von Grumbkow, 2007). Other studies have focused on emoticon use in laboratory settings in order to better understand the influence of emoticons on interpreting messages (Derks et al., 2007, Lo, 2008, Walther and D’Addario, 2001). By contrast, this is an examination of emoticon content within text messages obtained from users’ smartphones in the wild.

As such, we seek to holistically understand how people use emoticons via this ubiquitous and private method of communicating. Real communications data were collected from users’ iPhones unobtrusively over the course of 6 months. These data were analyzed to characterize the frequency and variety of emoticon usage through the SMS channel. We also build on previous research and assess gender differences.

Communication is not just a matter of speaking, writing, and interpreting words. Indeed, it is a complex process that involves factors such as content, language, grammar, experience, and nonverbal cues (Rezabek & Cochenour, 1998). Many researchers have noted the importance of nonverbal cues to understanding the meaning and nature of the message in F2F (Argyle, 1988). However, communication theories (e.g., social presence theory) have purported that CMC lacks contextual information and that the medium is disruptive for understanding the content and nature of messages (Sproull and Kiesler, 1986, Walther, 1992). For instance, this lack of contextual information has been blamed for causing electronic message recipients to perceive the senders of those messages as behaving rudely and offensively (Jenson, 2005).

More recent work, however, has shown that emoticons can provide this information and enhance CMC (Derks, Fischer, & Bos, 2008). Walther and D’Addario (2001) defined emoticons as graphic representations of facial expressions that are embedded in electronic messages. These often include punctuation marks and letters to create expressions such as happy, sad, or frustrated (which appear :), :(, and :/ respectively). Many researchers have suggested these cues enhance written communication in the same way visual or body language supports verbal communication (Derks et al., 2008, Rezabek and Cochenour, 1998). When studied empirically, viewing text online without emoticons led to incorrect interpretation of the nature of the message and the senders’ attitude (Lo, 2008). The inclusion of emoticons helped readers better understand the level and direction of the emotional context surrounding the message relayed over the internet.

Rezabek and Cochenour (1998) analyzed emails on listservs for emoticon content and frequency of use. Emoticons were used in 1–25% of the emails compiled from various listservs. According to the authors, many factors that could have influenced the large variance across emails were not assessed (e.g., social tie strength, gender, age, location, etc.). The listserv with the most messages (N = 349) consisted of only 6% that contained at least one emoticon.

Emoticons are used more often in synchronous communications (Derks et al., 2008). In instant messaging, positive emoticons were used at a higher frequency than negative emoticons and the use of emoticons strengthened the valence of the message (Lo, 2008). Emoticons were primarily used to express emotion, strengthen messages, and display humor or sarcasm. In studies of distributed teamwork, users regularly opted to use emoticons in team communications when such emoticon utility was made available. In situations where the teams did not have emoticon use available, users were not as satisfied with the system used to complete the artificial task in the laboratory (Rivera, Cooke, & Bauhs, 1996).

Because of the brevity of SMS communications and the fact that it is used both synchronously and asynchronously (Kasesniemi & Rautianen, 2002), it is possible that emoticon usage may have different patterns of use or enhanced importance. Ling (2005) examined 882 messages using phone interviews to gather data and found that only 6% of these messages contained emoticons. In other survey-based research (Qiao, 2010), 88% of a Chinese sample used emoticons. These users preferred SMS over other media and F2F communications to express emotions to others in their social networks. Findings also showed that these users largely used emoticons in SMS for humor and as a substitute for non-verbal communication.

Since females use more non-verbal communication in F2F encounters (Derks et al., 2008), researchers have been interested in understanding if this is also true in CMC. Interestingly, studies focused on gender differences have yielded mixed results. On the web (Wolf, 2000), males did not use many emoticons on sports newsgroups where most other viewers are also male. However, when males joined mixed-gender groups, they used emoticons more frequently. The authors suggested that both males and females sought to clarify emotional states when both genders were viewing their content. In addition, though the frequency of emoticons found on these mixed-gender forums were roughly equivalent, males and females used them differently. The former used them for humor and to display emotions while the latter used them mostly for sarcasm. A similar gender-use pattern has been found on IM (Lee, 2003).

Emoticon use for different kinds of tasks (task-oriented and social-oriented) has failed to show gender differences (Derks et al., 2007). While emoticons are used more in socially oriented tasks overall, males and females both use emoticons at the same rate. In another study (Huffaker & Calvert, 2005), content on blogs was examined for gender differences. Contrary to the popular typecast, male weblogs consisted of more emoticons compared to female weblogs.

Clearly, across all mediated communication methods, there are not static gender differences (Derks et al., 2008). Emoticons are used differently across tasks, contexts, and mediums. Similar to facial expressions and other non-verbal communications, emoticons are helpful to communicate social cues, emotion, and clarify the meaning of the message. How often they are really used in SMS as well as gender differences in patterns of usage are empirical questions we address in the present research.

Given the importance of emoticons in providing socioemotional context, we expected a large percentage of messages would contain at least one emoticon. The research on gender differences in emoticon use is mixed. However, many of the studies that found no differences either used contrived tasks or analyzed public content. Since text messages are mostly private communications with friends (Hakkila and Chatfield, 2005, Ling, 2005), we expected that our results would most closely resemble research found in F2F communications, with females sending more emoticons in their text messages compared to males. We also expected females to use a wider variety of emoticons compared to males.