In computer-mediated communication (CMC) users can purchase products, form
romantic relationships, or seek emotional support and so on. These interactions can
often occur between anonymous parties in one-off encounters. For instance, an eBay
buyer may purchase a product from a seller, never to meet the seller again. In the past,
research has focused on designing systems that will motivate members to act in a
trustworthy manner as well as understanding the conditions that lead a member to
violate another’s trust. Inevitably though, trust can break down; this previous research
does not consider the reciprocal emotional responses that occur as a result.
This thesis examines dyadic trust breakdowns that occur in online one-off, anonymous
interactions. A first aim of this work is to understand the emotional responses of the
offender and the victim of the trust breakdown. The view taken is that emotions involve
a reciprocal exchange in which the offender’s emotion, e.g. guilt, can elicit a response
from the victim, e.g. forgiveness. A second aim is driven by a reactive (as opposed to a
passive) design philosophy. Emotion theories are used as a basis to design online
mechanisms, which on a motivation level prompt the offender to repair the trust
breakdown, while at the same time they stimulate the victim to accept the offender’s
reparative gesture.
In particular, this thesis describes:
An argument, based on the relevant review of psychology, that lower self-awareness
fostered in certain online settings hinders an offender’s experience of shame, guilt or
embarrassment; and as a result of this, the offender is motivated to perpetuate the
offensive behaviour.
An experiment which substantiates these theoretical conclusions by providing a
mechanism that activates self-conscious emotions and in turn motivates the reversal
of offensive behaviour.

An argument, based on a substantive review of the psychology literature, that
forgiveness is a putative mechanism that repairs trust.
A formal framework for forgiveness which is used to construct an intelligent
forgiveness intervention mechanism which informs the victim of reasons for
forgiveness.
An experiment that hypothesised and showed that systems designed to stimulate
forgiveness, such as the forgiveness intervention mechanism, can restore a victim’s
trust in the offender.
This work contributes to the advancement of CMC on a theoretical level as well as to
the design of online systems. In the first instance, offensive behaviour is discussed
against a solid theoretical background on self-conscious emotions as well as theories of
CMC. This theory is then used to design mechanisms aiming to reverse users’ offensive
strategies. In a second step, the need to repair a trust breakdown is formulated on the
foundation of forgiveness theory. Prescriptions made by this theory are collectively used
to build and evaluate a mechanism that facilitates forgiveness. By using theory as a
foundation for designing proof of concepts, designers are provided with tangible
solutions for reversing and repairing a trust breakdown.