Reading Between the Lines, Analyzing Computer-Mediated Conversations using LIWC-22
Academic Level at Time of Presentation
Senior
Major
Psychology / Applied Behavior Analysis
Minor
Social & Behavioral Sciences
List all Project Mentors & Advisor(s)
Dr. Jana Hackathorn
Presentation Format
Oral Presentation
Abstract/Description
Computer-mediated interactions are said to have a communicative advantage, with more allowances for individuals to adjust how they present themselves, and produce an optimal presentation to others. (Walther, 1996). There are less nonverbal cues in CMC than in face-to-face (FTF) interactions, allowing more cognitive effort to be allocated to the tone and word choice used in message composition. This process of selective self-presentation has traditionally been studied through faux email correspondence. During composition, “editing” behaviors were coded and compared across conditions. (Walther, 2006). The purpose of this study is implementing modern technology for data collection, and increasing mundane realism. This study captured video data depicting participant’s typing behavior in response to a real-time interactive chat conversation with a confederate unknowingly, along with manipulated real-time positive/negative feedback presented onscreen dependent on condition assigned before arrival. Transcripts on participant conversations with confederates were also analyzed for a variety of factors, including affect, tone, and use of personal pronouns using the LIWC-22 tool. LIWC-22 offered analysis of each message of each participant, totaling more than 2000 data points. LIWC-22, despite its capabilities, also carries certain considerations for this use-case, and could ultimately fail if precautions in study design are not taken.
Spring Scholars Week 2026
Psychology: Completed Projects
Reading Between the Lines, Analyzing Computer-Mediated Conversations using LIWC-22
Computer-mediated interactions are said to have a communicative advantage, with more allowances for individuals to adjust how they present themselves, and produce an optimal presentation to others. (Walther, 1996). There are less nonverbal cues in CMC than in face-to-face (FTF) interactions, allowing more cognitive effort to be allocated to the tone and word choice used in message composition. This process of selective self-presentation has traditionally been studied through faux email correspondence. During composition, “editing” behaviors were coded and compared across conditions. (Walther, 2006). The purpose of this study is implementing modern technology for data collection, and increasing mundane realism. This study captured video data depicting participant’s typing behavior in response to a real-time interactive chat conversation with a confederate unknowingly, along with manipulated real-time positive/negative feedback presented onscreen dependent on condition assigned before arrival. Transcripts on participant conversations with confederates were also analyzed for a variety of factors, including affect, tone, and use of personal pronouns using the LIWC-22 tool. LIWC-22 offered analysis of each message of each participant, totaling more than 2000 data points. LIWC-22, despite its capabilities, also carries certain considerations for this use-case, and could ultimately fail if precautions in study design are not taken.