Krisela Rivera, Nancy J. Cooke, and Jeff A. Bauhs
New Mexico State University
Department of Psychology
Las Cruces, NM 88003
As technology advances, we are shifting from direct face-to-face or voice to voice interactions to computer-mediated communication (CMC). As a result of this shift the nature of communication has changed; in particular the ability to convey emotion is less straight forward. Twenty three subjects participated in a simulated, remote-CMC, group-decision making session. Twelve subjects had emoticons available, although use of these icons was optional. The remaining eleven did not have emoticons available. Dependent measures included user satisfaction, user frustration, conformity, length and focus of message, satisfaction with CMC system, and recall of communication events. The results indicated that subjects with emoticons used them and were more satisfied with the system than those subjects without emoticons. Thus it appears that users respond to emoticons and interpret them as intended.
Computer-mediated communication, groupware, computer-supported-cooperative work, distributed cognition, icons, emotions, and emoticons
As technology advances, communication is becoming increasingly indirect with e-mail and groupware replacing face-to-face or voice-to-voice communication. One type of technology finding its way into group decision-making settings is a form of computer software which facilitates computer-mediated-communication (CMC) (1). Although CMC can be incorporated in to face-to-face meetings, the focus of the current research is remote communication, which reduces the costs associated with convening group members for a meeting (5). These systems may facilitate in time management and information distribution. In fact, researchers have demonstrated that CMC enhances productivity in brainstorming tasks (4). Other research has shown that CMC equalizes group members' participation by allowing access regardless of social status (2,5). However, CMC systems are not without their problems. With the loss of more direct interaction (face-to-face or voice to voice), the communication process is altered. For example, users may have greater difficulty understanding computer-mediated "conversation" rather than face-to-face communications. Such misunderstandings may be due, in part, to a loss of information (7), including the emotional content of the message as indicated by facial expressions or voice fluctuations. Instead, emotional expressions are masked by the computer technology. The capability to express emotion in remote CMC situations may enhance decision making conducted via these systems. One method of transmitting such information is through the use of emoticons. For example, if a group member is happy with a decision, she might send a smile icon with her message, [ e.g. :-) ]. However, research has not examined the use of "emoticons" in remote CMC systems. The purpose of this research is to examine the effects of introducing emoticons into remote CMC situations.
Twenty-three subjects participated in a simulated remote CMC session. The session was simulated in that subjects believed they were interacting with three other group members, but in actuality the comments of these other group members were generated by the experimenter. Twelve of the subjects had six emoticon buttons available for their use (see Figure 1). Subjects selected these buttons to insert the desired emoticon into the text of the message. Use of these emoticon buttons was optional. The remaining 11 subjects did not have access to emoticon buttons. All subjects completed two different decision-making tasks: a selection task and a survival task. In the selection task, subjects were asked to work with the other group members to select the best candidate of three for an elementary school teacher position. In the survival task (3), subjects worked with the group to determine the most useful survival item from a set of eight items. Task order was counterbalanced across subjects. For each task the subjects first recorded their selection on an answer sheet (paper and pencil), then they received a computer tutorial on the CMC system. Next, they participated in a simulated remote CMC session in which they discussed the problem with the group. Subjects in both conditions received same can messages except in cases in which they asked specific questions which were then answered on-line. However, those subjects in the icon condition received messages which included emoticons. Following each task on the computer, subjects were asked to recall specific communication events. After completing both tasks, subjects filled out a questionnaire concerning the system. Dependent measures included user satisfaction, user frustration, conformity, length and focus of message, satisfaction with CMC system, and recall of communication events.
Figure 1. Emoticon buttons used in this study. In order to test icons currently created and used by many with word processors; icons were created using combinations of normal text characters.
Results indicated that when emoticons were available, they were used by 10 of the 12 subjects. Furthermore, 75% of the 12 subjects responded favorably when asked whether CMC enhances group-decision making, compared to 46% of the subjects in the no-emoticon condition, although this difference is only marginally significant (X 2 (1) = 2.10, p = .14).
In order to more closely examine the effects of emoticon availability on the communication process, the amount of decision conformity, the number of events recalled, and the length and types of statements made were analyzed in four separate 2 X 2 (emoticon availability by task) mixed ANOVAs. The analysis of the presence or absence of conformity (a decision change to correspond to the group majority) indicated that neither the availability of emoticons, nor their interaction with task, affected decision conformity. However, subjects tended to conform more in the survival task (61% changed) than the selection task (35% changed; F (1,21) = 4.13, p = .055).
A second ANOVA conducted on the number of events (0-4) correctly recalled, similarly indicated that emoticon availability and its interaction with task had no effect on event memory. However, there was a tendency for subjects to recall on average more events in the survival task (2.7) than in the selection task (2.3; F (1,21) = 3.39, p = .079).
In addition, emoticons and the emoticon by task interaction had no effect on the length of subjects' statements which were however, longer on average in the survival condition (131 words) than in the selection condition (109 words; F (1,21) = 4.12, p = .055). Interestingly, the presence of emoticons did affect the proportion of messages directed at the angry simulated member F (1,21) = 11.57, p = .0027. Subjects who had access to emoticons directed 44% of their messages (to the angry member in comparison to those without access to emoticons (20%).
Finally, subjects' opinions of the communication process itself, indicated that emoticon availability served to heighten complaints of lack of equality of participation (t (21) = 1.94, p = .066) and dissatisfaction with the overall communication experience (t (21) = 2.04, p = .054). These rather negative opinions of the communication process by emoticon users are interesting when coupled with the finding that emoticons users tended to more likely to respond favorably to the CMC system in general.
These results indicate that subjects used emoticons when available and that they were satisfied with the system more so than subjects who had no access to emoticons. Also, the presence of emoticons did not affect decision conformity, memory for communication events or message length, but it did affect the focus of the messages. In a previous study conducted by the authors when the ratio of positive to negative emoticons was 9 to 1 subjects talked to the simulated members in about equal proportions. However, in this study subjects with access to emoticons spent much of their time communicating with the group member who consistently displayed negative emoticons (angry, frustrated, confused), as opposed to the members who displayed positive emoticons (happy, surprise) the ratio of positive to negative emoticons was 1 to 1. Perhaps this explains why emoticon users were frustrated with the communication process. This focus on negative emoticons is not unlike face-to-face communication in which negative events have a higher impact compared to positive events (6). Users appear to attend to emoticons and interpret them as intended. Thus, an emoticon palette could be a useful tool for users trying to communicate emotional expressions via remote.