<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Evaluating Persuasive Systems Using the PSD Framework</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Ifeoma Adaji</string-name>
          <email>ita811@mail.usask.ca</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Julita Vassileva</string-name>
          <email>jiv@cs.usask.ca</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Table 2. Postulates of the PSD framework</string-name>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Personalizatio</string-name>
          <xref ref-type="aff" rid="aff5">5</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Self-</string-name>
          <xref ref-type="aff" rid="aff4">4</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Surface</string-name>
          <xref ref-type="aff" rid="aff2">2</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Real-world</string-name>
          <xref ref-type="aff" rid="aff3">3</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Department of Computer Science, University of Saskatchewan</institution>
          ,
          <addr-line>Saskatoon</addr-line>
          ,
          <country country="CA">Canada</country>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>Dialogue Social System, Support Support Credibility</institution>
          ,
          <addr-line>Support</addr-line>
        </aff>
        <aff id="aff2">
          <label>2</label>
          <institution>credibility</institution>
        </aff>
        <aff id="aff3">
          <label>3</label>
          <institution>feel</institution>
        </aff>
        <aff id="aff4">
          <label>4</label>
          <institution>monitoring</institution>
        </aff>
        <aff id="aff5">
          <label>5</label>
          <institution>n</institution>
        </aff>
      </contrib-group>
      <abstract>
        <p>The Persuasive Systems Design (PSD) is a framework for developing and evaluating persuasive systems. Despite its extensive use as a guide for developing persuasive systems, its use as an evaluation tool for persuasive systems is yet to be exploited. In this paper, we review the PSD framework and use it to evaluate the persuasiveness of Netflix. We were able to identify at least one implementation of all the persuasive principles of the PSD framework.</p>
      </abstract>
      <kwd-group>
        <kwd>Social learning</kwd>
        <kwd>Social comparison</kwd>
        <kwd>Normative influence</kwd>
        <kwd>Social facilitation</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>1. INTRODUCTION</title>
      <p>
        There is currently extensive research on the use of technology to
change people’s behavior and attitude without coercion, referred to
as persuasive technology [
        <xref ref-type="bibr" rid="ref1">1</xref>
        ]. These research include the use of
adaptive persuasive systems to increase the effects of persuasion,
such as [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ]. There is however limited research on how to evaluate
these persuasive systems. This paper reviews an evaluation
framework for persuasive systems; the Persuasive Systems Design
(PSD) framework. This framework comprises of persuasive
principles that could be used to develop and implement strategies
that encourage personalization and adaptation to user preferences.
We evaluated Netflix as a persuasive system using the PSD model.
2. PERSUASIVE SYSTEMS DESIGN (PSD)
      </p>
    </sec>
    <sec id="sec-2">
      <title>FRAMEWORK</title>
      <p>The PSD framework is a framework for designing an evaluating
persuasive systems [3]. It explains the content and functionalities
expected in a persuasive system and how a system’s design
principles can be converted to software requirements and then
implemented as features of the system. [3]. The PSD framework
comprises of three phases which should be considered in designing
and evaluating persuasive systems: 1) understanding key issues
behind persuasive systems, 2) analyzing the persuasion context and
3) design of system features.</p>
    </sec>
    <sec id="sec-3">
      <title>2.1 Understanding Key Issues Behind</title>
    </sec>
    <sec id="sec-4">
      <title>Persuasive Systems</title>
      <p>This phase of the PSD model describes seven postulates that should
be addressed in the design or evaluation of a persuasive system.
These postulates are listed in table 1.
Information technology is never neutral
People like their views about the world to be organized
and consistent
Direct and indirect routes are key persuasion strategies
5
6
7</p>
      <sec id="sec-4-1">
        <title>Persuasion is often incremental</title>
        <p>Persuasion through persuasive systems should always be
open
Persuasive systems should aim at un-obstructiveness
Persuasive systems should aim at being both useful and
easy to use
2.2</p>
      </sec>
    </sec>
    <sec id="sec-5">
      <title>Analyzing the Persuasion Context</title>
      <p>According to the PSD framework, for a system to be persuasive,
the persuasion process has to be analyzed in the use context;
understanding the role of the persuader, the persuadee, what
message is being communicated, the channel of communication
and in what context the message is being used. In addition,
persuasive systems should also analyze the user context. This
includes analyzing the user’s needs, interests, motivations, abilities,
persistence to change and personality among others.
2.3</p>
    </sec>
    <sec id="sec-6">
      <title>Design of System Features</title>
      <p>This phase of the PSD model describes the design principles that
should be considered when designing or evaluating persuasive
systems. There are 28 principles categorized into four groups based
on the task the principle is set to accomplish. The principles and
their respective categories are described in table 2.</p>
      <sec id="sec-6-1">
        <title>Rehearsal</title>
      </sec>
      <sec id="sec-6-2">
        <title>Social role</title>
      </sec>
      <sec id="sec-6-3">
        <title>Recognition</title>
      </sec>
    </sec>
    <sec id="sec-7">
      <title>3. Netflix As A Persuasive System</title>
      <p>Netflix is an acclaimed leading online television network that
provides TV shows and movies to its clients for a fee with over 81
million viewers in over 190 countries [4]. We chose Netflix as our
case study because of its popularity among users. Due to space
constraints, of the three phases of evaluating a persuasive system
using the PSD framework (described in section 2), we only
described our findings from the third phase, design of system
features. Of the four categories in this phase, we studied the first
two; primary task support and dialogue support categories. We
present these findings in the following section.</p>
    </sec>
    <sec id="sec-8">
      <title>3.1 Primary Task Support Category</title>
      <p>Principles in this category support users in carrying out their
primary tasks in a system. We were able to identify
implementations of all seven principles in this category. Table 3
shows the result of our review of Netflix using the persuasive
principles in this category.</p>
      <sec id="sec-8-1">
        <title>Persuasive principles and their definition Table 3. Analysis of Netflix using Primary Task Support</title>
      </sec>
      <sec id="sec-8-2">
        <title>Reduction</title>
        <p>Reducing a complex
task into smaller ones</p>
      </sec>
      <sec id="sec-8-3">
        <title>Tunneling</title>
        <p>Guiding a user through
a process while
persuading the user
along the way</p>
      </sec>
      <sec id="sec-8-4">
        <title>Tailoring</title>
        <p>Adapt information to
specific user groups</p>
      </sec>
      <sec id="sec-8-5">
        <title>Personalization</title>
        <p>Persuasive systems
should offer
personalized content to
individuals</p>
      </sec>
      <sec id="sec-8-6">
        <title>Self –monitoring</title>
        <p>Persuasive systems
should enable users to
monitor their
performance or
progress</p>
      </sec>
      <sec id="sec-8-7">
        <title>Simulation</title>
        <p>Persuasive systems
should allow users to
see the link between
cause and effect</p>
      </sec>
      <sec id="sec-8-8">
        <title>Rehearsal</title>
        <p>Persuasive systems
should allow users to
rehearse a target
behavior
Users can register and login with
their Facebook accounts instead of
filling out a registration form
Option to rate a movie after
watching with suggestions of what
to watch next
Users can tailor and personalize
movie suggestions to suit their needs
from their profile page. Hence
Netflix adapts to fit users individual
preferences and choices ensuring
that users with different movie
preferences do not receive the same
movie suggestions
Users can track what movies they
have viewed previously, their
viewing activity, ratings and
reviews they have written in the past
Netflix gives potential customers a
one month free trial period where
they get to use Netflix like paying
users. They have the option of
canceling this service without
penalty before the one month period
if they decide not to become paying
members</p>
      </sec>
    </sec>
    <sec id="sec-9">
      <title>3.2 Dialogue Support Category</title>
      <p>The principles in this category support dialogue between users and
the system which helps users achieve their goal or target behavior.
Table 4 describes these principles and examples of their
implementation in Netflix.</p>
      <sec id="sec-9-1">
        <title>Praise</title>
        <p>Users can praise reviews of other
users by voting them as helpful or
not.</p>
        <sec id="sec-9-1-1">
          <title>Systems that offer praise can be more persuasive than those that do not</title>
        </sec>
      </sec>
      <sec id="sec-9-2">
        <title>Rewards</title>
        <p>Persuasive systems
should offer rewards to
encourage users to keep
carrying out their target
behavior</p>
      </sec>
      <sec id="sec-9-3">
        <title>Reminders</title>
        <p>Systems that remind
users of their target
behavior are likely to be
more persuasive</p>
      </sec>
      <sec id="sec-9-4">
        <title>Suggestion</title>
        <p>Persuasive systems
should offer suggestions
to users</p>
      </sec>
      <sec id="sec-9-5">
        <title>Similarity</title>
        <p>A persuasive system
should behave similarly
to its users.</p>
      </sec>
      <sec id="sec-9-6">
        <title>Liking</title>
        <p>Systems that appeal to
the user are likely more
persuasive</p>
      </sec>
      <sec id="sec-9-7">
        <title>Social role</title>
        <p>Persuasive systems
should take up a social
role.</p>
        <sec id="sec-9-7-1">
          <title>Netflix rewards its new users with a free one month trial.</title>
        </sec>
        <sec id="sec-9-7-2">
          <title>Netflix emails its users when new</title>
          <p>movies are added to the system.
Netflix also reminds users to rate
and review a movie right after
watching to improve future
recommendations.</p>
          <p>Netflix offers movie suggestions to
users in various categories like
popular on Netflix, trending now,
Netflix originals, top picks for you.</p>
        </sec>
        <sec id="sec-9-7-3">
          <title>Users can create and personalize their profiles so that movie recommendations are as close to the user’s preferences as possible.</title>
        </sec>
        <sec id="sec-9-7-4">
          <title>The appeal of Netflix to users is</title>
          <p>subjective and can only be
determined through a user study,
but it`s high customer base could
be an indication of its appeal.
Netflix has a life chat session to
support users that need help.</p>
        </sec>
      </sec>
    </sec>
    <sec id="sec-10">
      <title>4. CONCLUSION</title>
      <p>Though persuasive technology is widely used, there has been little
research on evaluating persuasive systems. This paper reviews the
Persuasive Systems Design (PSD) framework, a framework for
developing and evaluating persuasive systems. This framework
comprises of persuasive principles that could be used to develop
and implement strategies that encourage personalization and
adaptation to user preferences. Using Netflix as a case study, we
identified the implementation of the persuasive principles in the
third phase of the mode; design of system features. This study can
act as a guide for the development of evaluation metrics for
persuasive systems. A user study to validate the effectiveness of
these principles is currently being proposed.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          [1]
          <string-name>
            <given-names>I.</given-names>
            <surname>Wiafe</surname>
          </string-name>
          and
          <string-name>
            <given-names>K.</given-names>
            <surname>Nakata</surname>
          </string-name>
          , “
          <article-title>Bibliographic analysis of persuasive systems: techniques, methods and domains of application,” Persuasive</article-title>
          . Technology, vol.
          <volume>61</volume>
          ,
          <year>2012</year>
          .
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          [2] [3]
          <string-name>
            <given-names>M.</given-names>
            <surname>Kaptein</surname>
          </string-name>
          , B. De Ruyter,
          <string-name>
            <given-names>P.</given-names>
            <surname>Markopoulos</surname>
          </string-name>
          , and
          <string-name>
            <given-names>E.</given-names>
            <surname>Aarts</surname>
          </string-name>
          ,
          <article-title>Adaptive persuasive systems: a study of tailored persuasive text messages to reduce snacking</article-title>
          .
          <source>ACM Transactions on Interactive Intelligent Systems (TiiS)</source>
          , vol.
          <volume>2</volume>
          , pp.
          <fpage>10</fpage>
          ,
          <year>2012</year>
          .
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          <string-name>
            <given-names>H.</given-names>
            <surname>Oinas-Kukkonen</surname>
          </string-name>
          and
          <string-name>
            <given-names>M.</given-names>
            <surname>Harjumaa</surname>
          </string-name>
          , “
          <article-title>A systematic framework for designing and evaluating persuasive systems</article-title>
          ,” in Persuasive technology, Springer,
          <year>2008</year>
          , pp.
          <fpage>164</fpage>
          -
          <lpage>176</lpage>
          .
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>