Download Personalized Persuasion in Ambient Intelligence - Maurits Kaptein PDF

TitlePersonalized Persuasion in Ambient Intelligence - Maurits Kaptein
File Size5.2 MB
Total Pages234
Table of Contents
	Background: Persuasion in Intelligent Systems
	Research Problem
	Key Contributions
Related Work
	Persuasive Technologies
	Attitude and Behavioral Change
	Conclusion: Reflections
Insight Generation I: Heterogeneity in Responses to Social Influence Strategies
	Study 1: Identifying an upper bound in heterogeneity
	Study 2: Examining stability of heterogeneity over time
	Study 3: Stability of heterogeneity across context
Insight Generation II: Choice and Repetition Effects of Influence Strategies
	Study 1: Choice Effects
	Study 2 and 3: Simultaneous Presentation
	Study 4: Simultaneous Presentation in Practice
Insight Generation III: Meta-Judgemental Measures
	Introduction: Measures of Persuasion Susceptibility
	Study 1: Initial application of the short STPS
	Study 2: Development and Validation of the STPS
Design Requirements
	Conclusions From the Insight Generation Chapters
	Identification, Representation, and Measurement
	Dynamic Adaptation
	Properties of Adaptive Persuasive Systems
Case Studies I: Influencing Health Decisions
	Introduction: Measures of Susceptibility
	Study 1 & 2: Promoting Lunch Walks
	Study 3: Influencing Snacking Behavior
Case Studies II: Dynamically Personalizing Persuasion
	Introduction: Dynamic Persuasive Systems
	Design 1: APStairs, Promoting Stair Usage
	Design 2: Influencing Docking Behavior
	Design 3: E-Selling Applications
	Advancing the Design of Persuasive Technologies
	The Possible Future(s) of Persuasion Profiles
	Ethical Considerations
	Future Research Challenges
Articles by Maurits Kaptein
			List of Figures
			List of Tables
Document Text Contents
Page 1

Personalized Persuasion

Ambient Intelligence

Page 2

This research was conducted in part at the User Centered Engineering
group, Department of Industrial Design, Eindhoven University of
Technology (Eindhoven, The Netherlands), in part at Philips Research
Laboratories (Eindhoven, The Netherlands), and in part at the
Department of Communication, Stanford University (Stanford, CA).
This research was commissioned and funded by Philips Research.

An electronic copy of this thesis in PDF format is available from
the TU/e library website (

2011-2012, Maurits Clemens Kaptein, The Netherlands
All rights reserved. Reproduction of this publication in whole or in
part is prohibited without the prior permission from the author.

A catalogue record is available from the Eindhoven University of Technology
ISBN: 978-90-386-3106-6

Page 117


6.4 Properties of Adaptive Persuasive Systems
In this chapter the results of the three insight generation chapters were
summarized. These results motivate the need for designers to build
adaptive persuasive systems: Systems that use persuasion profiles to
adapt to responses of individual users to increase their effectiveness.
The next two Chapters (7 and 8) explore the effects of usage of both
static (Chapter 7) and dynamic (Chapter 8) persuasion profiles in per-
suasive systems. Before presenting these case studies it is important
to make clear what makes persuasion profiles—and thus profiles that
describe people’s susceptibility to different means to an end—distinct
from other profiles that identify people preferred ends—such as those
used by recommender systems.

6.4.1 End-independence
Means-adaptive persuasive technologies—technologies that use persua-
sion profiles—are distinctive from other adaptive systems such as rec-
ommender systems in their end-independence: a persuasion profile cre-
ated in one context can be applied to bringing about other ends in
that same context or to behavior or attitude change in a quite differ-
ent context. This feature of persuasion profiling is best illustrated by
contrast with end adaptation. Any adaptation that selects the particu-
lar end (or goal) of a persuasive attempt is inherently context-specific.
Though there may be associations between individual differences across
context (e.g., between book preferences and political attitudes) these
associations are themselves specific to pairs of contexts. On the other
hand, persuasion profiles are designed and expected to be independent
of particular ends and contexts.

It is important to clarify exactly what is required for end-
independence to be obtained. If a persuasion profile is end-independent
than this does not imply that the effectiveness of all social influence
influence strategies for a specific individual is constant across all con-
texts. Consistent with the results reviewed in Chapter 2, it is clear that
influence strategy effectiveness depends on, e.g., the type of behavior
change or the involvement of users. That is, the most effective influ-
ence strategy for a system to employ, even given the user’s persuasion
profile, would depend on both context and target behavior. However,
end-independence requires only that the difference between the average
effect of a strategy for the population and the effect of that strategy for
a specific individual is relatively consistent across contexts and ends.

Page 118


6.4.2 Non-disclosure
Designers of means-adaptive persuasive systems can be tempted not to
disclose the adaptation of their system. This can be contrasted with
end adaptation, in which it is often advantageous for the persuader to
disclose the adaption and the adaptation is (potentially) easy to detect.
For example, when recommends books for an individual
it makes clear that these are personalized recommendations—thus ben-
efiting from effects of apparent personalization and enabling presenting
reasons why these books were recommended (E.g. Ochi et al., 2010). In
contrast, with means adaptation, not only may the results of the adap-
tation be less visible to users (e.g. emphasizing either “Pulitzer Prize
winning” or “International bestseller” ), but disclosure of the adaptation
may reduce the target’s attitude or behavior change.

It is hypothesized that the effectiveness of social influence strate-
gies is, at least partly, caused by automatic processes. According to
dual-process models (Cacioppo et al., 1986), under low elaboration mes-
sage variables manipulated in the selection of influence strategies lead
to compliance without much thought. Disclosure of means adaptation
can increase elaboration. This in turn might decrease the effectiveness
of influence strategies which operate primarily via the peripheral route.
More generally, disclosure of means adaptation is a disclosure of persua-
sive intent, which can increase elaboration and resistance to persuasion
(Cf. Tormala and Petty, 2004).

6.4.3 Conclusions
Based on the empirical results presented in the insight generation chap-
ters it is clear that persuasive technologies would benefit from personal-
ization of the selection of influence strategies. The ambient intelligence
scenario provides designers with the opportunity to identify, present,
and measure the effects of different influence strategies and is thus par-
ticularly suited for the design of such systems. This chapter presented
persuasion profilesas a means for designers to create systems that per-
sonalize their persuasive attempts.

The next two chapters will explore the use of persuasion profiles in
actual persuasive interventions. These case studychapters each present
multiple persuasive systems that are personalized—or adapted—to in-
dividuals. The first case study chapter, Chapter 7, focuses on the use of
static persuasion profiles obtained using meta-judgmental measures to
optimize several health interventions. The second case study chapter,
Chapter 7, presents three persuasive systems that use dynamic persua-

Similer Documents