Document Type


Date of Degree

Spring 2016

Degree Name

PhD (Doctor of Philosophy)

Degree In

Communication Studies

First Advisor

Andrew C. High


This dissertation examines the “buffer effect,” an important but understudied feature of computer-mediated communication (CMC). Research on the buffer effect posits that CMC venues provide a buffering “screen” that users can literally and figuratively hide behind. The buffer can make people feel more comfortable during interactions, and is theorized to be especially relevant in contexts where self-presentation is threatened. This study employs transgressions as ideal sites for examining the buffer effect because of the high level of threatened self-presentation involved therein.

The current project tests whether people perceive different levels of a buffer in different channels of communication, and how the buffer effect is related to other widely studied features of CMC, such as interactivity, synchronicity, and social presence. It also tests outcomes of the buffer effect for both senders and receivers of transgressive messages. Specifically, it posits that the buffer effect is beneficial to senders of transgressive messages, and is detrimental to receivers of those messages. Furthermore, in the context of transgressions, the amount of responsibility that a person takes for the transgression is a factor that influences how others perceive the situation. Therefore, the current study also considers receivers' perceptions of the level of responsibility the sender accepts, and specifically posits that senders' higher levels of responsibility are associated with positive outcomes for receivers.

The dissertation is comprised of two studies. In Study One, participants responded to a survey to test their perceptions of the buffer effect and of other features of CMC in various channels. Participants also responded to a hypothetical situation to indicate how the buffer effect influences outcomes when sending a transgressive message. Study Two employed an experimental procedure to test how senders and receivers perceive the buffer effect in actual interactions, as well as how senders' acceptance of responsibility affects outcomes for receivers. Half of the participants were assigned the role of sender and were trained to provide a transgressive message to the receiver. Specifically, senders were trained to say that they had to leave the experiment early without completing the study, rendering the receiver ineligible for course credit. Both the channel (i.e., face-to-face, instant messaging, text messaging) and the senders' level of responsibility (i.e., low/high) were manipulated.

Results suggested that the buffer effect manifests in different levels for various channels of communication, such that face-to-face environments provide the lowest buffer, followed by video chat, social networking sites, instant messaging, and email. Text messaging provides the highest buffer. The buffer effect is negatively related to other features of CMC (i.e., synchronicity, interactivity, and social presence) for low-buffer channels, and is either positively or not significantly related to these features in high-buffer channels. Results also suggest that the buffer effect is associated with benefits for senders in both hypothetical and actual interactions, but does not affect receivers' outcomes. Receivers' perceptions of the level of responsibility that senders accept affects receivers' outcomes, but only within environments with a low and moderate buffer. These results extend research on CMC and on transgressive communication. Results also offer practical implications for how people might elect to use channels and modify the content of their message when communicating a transgression to a friend.

Public Abstract

This dissertation explores the buffer effect, which is the literal and figurative shield that mediated environments provide. The buffer can potentially make people feel more comfortable when they have to admit that they have done something wrong, or committed a transgression. This project tests what outcomes the buffer effect produces for both partners in a conversation about transgressions.

Participants in Study One completed a survey to test their perceptions of the buffer effect and of other commonly studied features of mediated channels. They also indicated how they would respond to using channels with different levels of the buffer effect to admit a transgression to a friend. Participants in Study Two completed an experiment in which one participant was trained to tell the other that they could not complete the study as planned. Both the channel in which the partners communicated, as well as the level of responsibility that the message sender took, were manipulated. Both participants then indicated how they perceived the interaction.

Results suggested that different channels provide different levels of a buffer, and that the buffer effect is associated with other features of mediated channels. Senders of transgressive messages experienced more favorable outcomes in high-buffer channels, whereas the buffer effect did not influence receivers’ outcomes. However, when senders took more responsibility for the transgression, receivers experienced more positive outcomes than when the sender took less responsibility. Results provide practical implications for how people select different channels and modify their message when communicating a transgression to a friend.


publicabstract, Buffer Effect, CMC, Communication, Computer Mediated, Technology, Transgressions


xii, 229




Copyright 2016 Crystal DeAnn Wotipka

Included in

Communication Commons