Introduction
Technology is everywhere. Our homes and workplaces are filled with technology and
people themselves carry more and more technology with them everywhere they go.
Furthermore, persuasion is everywhere. From doing grocery shopping to reading the
news online, we are being persuaded to buy certain products or behave a certain way.
This persuasion is not new, only the way it is presented to us, is changing. With the
ubiquitous presence of technology, there is also ubiquitous presence of persuasion
through technology. However, because of the sometimes overwhelming possibilities
and instances of persuasion, it is even more important to select the right strategies to
actually influence the behavior of people and not just add to the ‘noise’ of persuasion
all around us. This poses real challenges for researchers and practitioners in the field
of Behavior Change Support Systems (BCSSs) [
1
].
The aim of a BCSS is to help people to change their behavior through technology.
Two domains that gave gained considerable attention are the health and the energy
domain. In the health domain, BCSSs have been developed to support people to
become more physically active, or to shop and eat healthier (e.g. [
2,3
]). In the energy
sector, BCSSs have been created to help people become aware of their energy
consumption and to reduce this (e.g. [
4
]). However, BCSSs also hold potential for other
domains, e.g. in the context of the workplace [
5,6
]. BCSSs may support people to
work according to protocols or to help them work more efficiently. In all these cases,
the BCSS helps people to change their behavior, but there is always a certain level of
motivation to start with. It is the challenge for BCSSs to find the right way and the
right moment to persuade the user to reach previously agreed upon goals.
Finding this right way of persuasion is a challenge that can be seen in the
development and evaluation of many BCSSs, but has not always received very structured
attention. There are multiple ways how to deal with this challenge. For instance, the
Persuasive System Design Model (PSD-model) [
7
] provides guidelines for developing
and evaluating BCSSs. It states that for developing BCSSs, the starting point is
understanding the postulates that underline each persuasive system. After that, one should
analyze the intent, event and strategy. Only then can developers decide on which
persuasive features can be used to reach the goals of the BCSS. The persuasive features
in the PSD-model are divided into four categories. Primary task support holds
strategies that are aimed to make the main task of the BCSS easier for the user. Examples
of strategies in this category are reduction (reducing complex behavior to simpler
tasks) and tailoring (presenting information that is tailored to the needs and wishes of
a target group). Strategies from primary task support seem to be used most in many
BCSSs, from web-based interventions to mobile apps [
8,9
]. This emphasis on primary
task support features seems logical from the perspective of the developers (i.e. most
of the time the BCSS is built to support this primary task) and many individual
features been shown to be effective in behavior change (e.g. the positive effects of
tailoring are well known [
10
]. However, deciding which primary task support features to
choose in which situation, is still a challenge. Moreover, primary task support is not
the only category of persuasive features and research shows that other categories
might even be more important for the persuasive power of BCSSs than primary task
support [
8,11,12
].
Dialogue support, the second category of persuasive features in the PSD-model, is
aimed at supporting the dialogue between the user and the system. Examples of
features from this category are reminders, praise and a social role. Reminders are a
feature that is well investigated and can play an important role in persuading users
[
13,14
]. Not surprising, this feature has been used in many BCSSs (see e.g. [
8,9
].
Other features form this category have been used less often, although with the recent
rise of gamification techniques, features as (virtual) rewards are getting more
attention (see e.g. [
15
]). Research shows that dialogue support as a category seems to pay
an important role in persuading users [
8,11,12
].
Credibility support, or the ways to make a system more credible and this persuasive,
has been used to some extent in BCSS [
9,16,17
]. Credibility is an important issue in
our digital community, but selecting the right way to support credibility of BCSS
seems to have gotten only limited attention.
The last category of persuasive features in the PSD-model is social support, or the
ways to design the system so that it motivates users by leveraging social influence.
Social support through technology has received a lot research attention (see e.g.
[1820]) and seems to provide huge opportunities to persuade people in our social
environment, which takes place more and more in our online world. The humanization of
the web [
21
] plays an important role in this respect. In this light, is it even more
surprising that social support features are used so little in BCSSs [
8,9,16,17
].
2
Advances in BCSS research
The Third International Workshop on Behavior Change Support Systems provides a
place to discuss recent advances in BCSS research. The six papers presented at the
workshop all cover the timely topic of how to select and evaluate effective persuasive
strategies. The first session includes three papers that are about selecting persuasive
strategies to increase the reach of BCSS. The second session covers three papers that
show the importance of the context of the BCSS.
2.1
Persuasive strategies to increase reach
Triggers in the environment; increasing the reach of Behavior Change Support
Systems by connecting to the offline worldby Ludden and Offringa [
22
] discusses why
many BCSSs have limited reach and provides an alternative approach to increase
reach; design for stages of change. Through a case study, the authors show that
triggers in the environment can be designed that may reach a larger group than an online
BCSS. By connecting these offline triggers to the BCSS, reach may be increased.
Enhancing persuasive features of Behavior Change Support Systems: the role of
UFADE by Wiafe and Frempong [
23
] builds on earlier work on the three-dimensional
relationships between attitude and behavior (3D-RAB) model [
24
] and addresses the
issue of changing user needs during the use of BCSSs. When BCSSs are not ready for
these changing needs, they become obsolete for a group of users, thereby diminishing
its reach. The Unified Framework for Analyzing, Designing and Evaluating
persuasive systems (U-FADE) provides an approach to evaluate and select persuasive
features based on the (changing) needs of the targeted users. Application to an existing
weight management intervention shows the possibilities of this approach.
How Persuasive are Serious Games, Social Media and mHealth Technologies for
Vulnerable Young Adults? Design Factors for Health Behavior and Lifestyle Change
Support: Sexual Health Caseby Kulyk, Den Haas, David and Van Gemert-Pijnen
[
25
] aims to identify effective persuasive features for a specific target group which is
not reached by regular interventions. They use the case of improving the sexual health
of vulnerable young adults and employ focus groups to gain insight into the needs and
attitudes of these young adults towards persuasive features and design factors that
contribute to the use and uptake of existing and new health technologies.
2.2
Importance of context
About the persuasion context for BCSSs: analyzing the contextual factorbsy Halttu,
Oduor, Tikka and Oinas-Kukkonen [
26
] emphasizes the importance of taking the
context into account for designing and evaluating BCSSs. They present the event
model, which is a conceptual tool for identifying which contextual factors are
important and how to take these factors into account. This event model is an extension
of the persuasion context of the PSD-model and provides more concrete factors that
potentially play a role in the everyday lives of end-users and should be taken into
account when investigating a BCSS.
Behavior Change Support Systems for Privacy and Securityby Kegel and Wieringa
[
27
] proposes to use a BCSS for a specific domain, namely information security. For
this specific context, the challenges seem to differ from areas that have received more
attention in the BCSS-field like healthcare and energy consumption. Challenges that
have been identified are motivation and change type, and the dynamic nature of
security threats. The authors propose the Personal Information Security Assistant (PISA)
as a possible solution to these challenges.
Persuasive information security: a Behavior Change Support System to help
employees protect organizational information securityby Busch, Patil, Regal, Hochleitner,
Fröhlich and Tscheligi [
28
] also tackles the issue of information security, but in the
context of the workplace. They present a Behavior Change Support prototype that
implements persuasive features (points, quiz, challenges, statistics, personalization
and risk communication) to support employees to comply with information security
policies. Preliminary results of user studies show that this is a promising approach for
influencing user attitudes and behaviors regarding secure work practices.
3