Dear CoC faculty members, staff, and fellow students:

You are cordially invited to my Ph.D. thesis defense, which will take place on Monday, February 23, 2026, from 3:00 PM to 5:00 PM (EST).

Title: Designing Relay: Towards Computer-Mediated Metacommunication

Stanley J. Cantrell

Ph.D. Candidate in Human-Centered Computing

School of Interactive Computing

Georgia Institute of Technology

Date: Monday, February 23, 2026

Time: 3:00 PM to 5:00 PM (EST)

Location: Price Gilbert 4222 (Georgia Tech Library Dissertation Defense Room)

Virtual Meeting:  Zoom

Meeting ID: 971 6856 0462

Passcode: relay

Abstract:

Computer-mediated communication (CMC) has become essential to how people sustain relationships, coordinate daily life, and express emotion across distance. Text-based communication platforms like Apple iMessage have emerged as a dominant communication channel amongst young adults. Yet the nonverbal cues that guide interpretation in face-to-face interactions, such as tone of voice, facial expression, and body language, are often attenuated or entirely absent in digital exchanges. Without these signals, text-messaging is reduced to the mere transmission of words, lacking the expressive depth and nuance that makes communication meaningful. This dissertation establishes computer-mediated metacommunication, defined here as “the signals that shape how a digital message is interpreted,” as both a critical area of inquiry within the HCI and CMC communities and as a foundational design goal for text-based communication platforms.

Early studies in this dissertation reveal that common metacommunicative cues are neither universally used nor universally understood by young adults. A cross-cultural examination of emoji and emoticon use shows that usage patterns vary by sex, race, and ethnicity, influencing how tone and intent are expressed and inferred. A subsequent study explores the viability of sonification as a means of making affective cues, such as Facebook Reactions, universally accessible to all technology users regardless of their level of visual acuity. Findings show that some auditory mappings produce reliably interpretable sonified Reactions while others expose where similar affective categories become sonically indistinguishable.

These insights motivate the introduction of Relay, a text-based communication platform that surfaces real-time feedback about tone and sentiment during digital conversations. Relay introduces novel metacommunicative features such as an emoji dictionary, sentiment indicators, and sentiment tagging designed to help users interpret incoming messages and reflect on how their own messages might be perceived. A concept evaluation examines how participants interpret these features and reveals key tradeoffs around clarity, user control, and when system-provided cues feel helpful rather than intrusive.

A deployment study extends this work by examining Relay in intimate, dyadic relationships over a multi-week period, leveraging these contexts as environments where emotional expression is rich and abundant, and where miscommunication can be observed as it naturally occurs. Findings show how Relay's features can draw attention to tone, encourage clarification, and support repair when ambiguity arises. The deployment also reveals that metacommunicative scaffolding is not uniformly helpful, and its value depends on relationship dynamics, communication goals, and individual receptiveness to system-generated metacommunicative feedback.

Overall, these contributions demonstrate the need for more intentional design of CMC systems that prioritize metacommunication. This dissertation explores multiple approaches through which this can occur, each revealing varying benefits, limitations, and guidance for designing effective computer-mediated metacommunication platforms. Relay demonstrated the advantages of metacommunicative features, showing that when platforms surface cues about tone, sentiment, and emotion, users become more attentive to how their messages might be received and more equipped to clarify meaning before misunderstandings escalate. As text-based communication continues to mediate how people connect, platforms must evolve beyond simply transmitting messages towards actively supporting the interpretive work that makes understanding possible.