Early Experiences with Participatory Design
of Ambient Persuasive Technology
Janet Davis
Grinnell College
Grinnell, IA 50112
davisjan@cs.grinnell.edu
November 3, 2008
1 Background
In the face of global warming and other environmental concerns, Grinnell College and many other institu-
tions want to reduce their environmental footprints. Researchers in computer science, too, are concerned
with technology’s impacts on the environment. For example, just as we insulate buildings to retain heat
and install low-flow showerheads, researchers are trying to decrease energy consumption of computer sys-
tems [17]. However, increasing efficiency to reduce resource consumption only goes so far. For further
resource conservation, changes in behavior are necessary.
Environmental conservation has been a significant domain in the new research area of persuasive tech-
nology: “interactive computing systems designed to change people’s attitudes and behaviors” [7, p. 1]. In
his seminal 1998 paper, B.J. Fogg identifies environmental sustainability as a key domain for persusasive
computing, using paper recycling as an example to illustrate different roles of persuasive technology [6].
While a number of researchers have developed web-based and mobile tools to to promote environmentally
sensitive choices, such as the SmartTrip errand-planning tool [10], the GreenScanner system for product
environmental impact reviews [18], and the Footprints social network tool [12], embedding devices in the
physical world has also proved a fruitful approach for providing feedback and information to households
and communities.
Several persuasive technology design efforts are intended to make resource consumption visible at the
time and place where resources are used. Indeed, researchers in environmental science have shown that
feedback, in combination with goal-setting [3, 19] or social comparison [16], is effective for reducing energy
consumption by 5–20% [5]. Feedback is most effective when it is immediate, at the time of energy use [5].
The goals of ambient displays—to provide awareness through the physical environment, without demanding
attention [20]—seem particularly suitable to providing feedback or reminders in the built environment. For
example, the power-aware cord uses animated LEDs embedded in a power strip to show at a glance how
much power is being drawn [8]. WaterBot aims to reduce water consumption by tracking and displaying
information about water use at the sink itself [1]. Intriguingly, Holstius, et al., use live and robotic plants in
an ambient display that shows the balance between trash and recycling in a dining area [9]. These efforts
show the area of ambient persuasive technology is ripe for further investigation.
While there has been much work on strategies for persuasion and on design of specific technologies,
there has been relatively little attention to methods for designing persuasive technology. Challenges include
1

bringing a user-centered rather than designer- or technology-centered focus to the design process [11], antic-
ipating unintended consequences of the persuasive technology [2], designing for user acceptance and even
ownership, and addressing ethical issues of consent and free will inherent in attempts to persuade [2], which
may be amplified by computer systems’ persistence and lack of moral judgment [7].
In response to this concern about ethical design methods, I am exploring the use of participatory design
methods to develop persuasive technology systems for promoting environmental conservation, tailored to
a specific group of stakeholders and context of use. Participatory design (or PD) is a family of theories
and methods that involve potential users as full participants in design processes leading to the development
of computer systems and computer-based activities [14]. Participatory design methods include workshops,
storytelling, role-playing, games, making low-tech models, and cooperative prototyping. Many, though not
all, PD researchers and practitioners are motivated in part by a belief in democratizing technology design.
Other motivations for using PD methods include mutual understanding on the part of users and designers,
user engagement in the design process and ownership of the eventual product, and the development of new
ideas drawing on multiple perspectives.
To the best of my knowledge, no prior work has considered participatory design of persuasive tech-
nology. Yet participatory design methods show promise in addressing challenges of designing persuasive
technology. A commonly cited advantage of participatory design is that it promotes a sense of ownership
among the technology users. The design process further benefits from participants’ creativity and their
knowledge, both explicit and tacit, about the context for technology use. Beyond these usual reasons to
employ participatory design, involving potential users in design helps to avoid some potential ethical issues
with persuasive technology. Without participatory design, the designer stands outside of the community and
intends to change the behavior of community members. With participatory design, the persuasive intent
comes in part from community members who want to change the community’s behavior from within.
2 Current work
In summer 2008, I worked with two undergraduate students, Timothy Miller and Patrick Rich, to engage
Grinnell students as partners in design of a persuasive system to reduce resource consumption in a public
location on campus [13]. We used a game-like technique to help participants look for places where resources
are consumed, introduced them to sensor and actuator technology through the Phidgets rapid prototyping
framework (www.phidgets.com), engaged them in building mockups, and invited them to critique early
prototypes of several possible designs.
Ultimately, we designed an interactive sculpture to occupy a staircase in the science building, intended
to attract people from a nearby elevator to the stairs [13]. The system uses LEDs embedded in hand-sized
wire sculptures to evoke fireflies along the walls and railings of the stairs. Some fireflies are equipped with
vibration sensors; tapping one of these fireflies is considered a “catch” and triggers a celebratory animation.
By pressing a button at the bottom or top of the stairs, the user can race a simulated elevator: the fireflies
sequentially light up at the speed at which the elevator travels, showing that most people can walk faster than
the elevator. When the elevator call button is pressed, a smaller group of fireflies near the elevator animate
to point in the direction of the stairs. As a whole, this temporary installation is intended to draw attention to
information posted nearby about the energy use of this particular elevator.
We are currently gathering baseline data on stair and elevator use and preparing to install the system. We
plan to evaluate the system with respect to frequency of stair and elevator use, frequency of interactions with
the system itself, and self-reports regarding the system’s effectiveness in promoting behavior and attitude
change.
2