To Be or Not To Be Emotional: Impression Formation Effects of Emoticons in Moderated Chatrooms
Student Researchers

Corina Constantin, Sriram Kalyanaraman, Carmen Stavrositu, Nathan Wagoner
This paper is based on a project from a graduate research methods course.

Faculty Supervisor

Dr. S. Shyam Sundar

For a complete report of this research, see:

Constantin, C., Kalyanaraman, S., Stavrositu, C., & Wagoner, N. (2002, August). To be or not to be emotional: Impression formation effects of emoticons in moderated chatrooms. Paper presented to the Communication Technology and Policy Division at the 85th annual convention of the Association for Education in Journalism and Mass Communication (AEJMC), Miami Beach, FL. [Top-three student paper award]


Early research on Computer Mediated Communication (CMC) has consistently shown that, due to reduced interactivity and social information, CMC is asynchronous and impersonal, suited mostly for task-oriented activities. Nowadays, however, CMC is becoming increasingly synchronous (due to such instant communication exchange applications as chat rooms or virtual communities), and its social orientation is growing more salient. Internet users are devoting increasing interest to online social interactions. Moreover, CMC communications have established a culture of their own, characterized by abbreviations, acronyms, spelling variations and emoticons. Starting from the premise that all these graphic cues have the potential of substituting the lack of Face-to-Face (FtF) communication tools, the present paper attempts to gauge impression formation effects of emoticons in a moderated chat room scenario by examining the following hypothesis:


H1: In a CMC context, a chat room moderator is more likely to elicit positive impressions in chat transcripts’ readers when he/she uses emoticons than when he/she does not.


A between-subjects factorial experiment was conducted with the participation of 58 undergraduate students. Use of emoticons was manipulated by creating four experimental conditions, as follows: just moderator with emoticons, just chatters with emoticons, everybody with emoticons, and none with emoticons. Subsequently, two chat transcripts were created: the San Francisco Chat (the moderator leads a discussion with the chatters about San Francisco and the Bay Area) and Reality vs. Imagination Chat (the moderator and the chatters talk about the ease/difficulty of making the distinction between reality and imagination), to reflect the four experimental conditions. Participants were evenly spread across the four conditions of the two chat transcripts. After exposure to the chat transcripts, impressions of the moderators were elicited via measures included in a paper-and-pencil questionnaire.


The dependent measure of impression formation was divided into six factors called “Conscientiousness,” “Boldness,” “Intensity,” “Professionalism,” “Dynamism” and “Confidence.”

H1: Not supported. On the whole, subjects did not evince positive impressions of moderators using emoticons. Significant main effects of moderator’s use of emoticons were found, suggesting that the moderator was perceived as more dynamic, friendlier, more valuable and more talkative when she/he did not use emoticons than when she/he did. Also, the moderator was perceived as more conscious, more sociable and nicer when she/he did not use emoticons.


There are at least three possible explanations for the lack of support for our hypothesis. One of them refers to a “negativity bias,” whereby if an initial negative impression of the moderator occurred supplemental nonverbal cues (in our case, emoticons) would emphasize such negative impressions. The second explanation, based on the model of mediated-communication competence, suggests that emoticons’ use might make a moderator appear too friendly, too familiar, not authoritative or expert enough to be taken seriously, resulting in moderator competency being questioned. Finally, the distinction between participant and observer in a communicational environment is worth noting - our participants did not take part in the chat conversation, such that their ratings of the moderators were based on observation, not participation.

For more details regarding the study contact

Dr. S. Shyam Sundar by e-mail at or by telephone at (814) 865-2173

>> Return to abstract listings page
>> Return to the main research page

Media Effects Research Lab at College of Communications, Penn State University