Content uploaded by Oladapo Oyebode
Author content
All content in this area was uploaded by Oladapo Oyebode on Mar 29, 2024
Content may be subject to copyright.
User Modeling and User-Adapted Interaction
https://doi.org/10.1007/s11257-023-09390-x
Persuasive strategies and emotional states:
towards designing personalized and emotion-adaptive
persuasive systems
Oladapo Oyebode1·Darren Steeves2·Rita Orji1
Received: 11 November 2022 / Accepted in revised form: 26 November 2023
© The Author(s), under exclusive licence to Springer Nature B.V. 2024
Abstract
Persuasive strategies have been widely operationalized in systems or applications to
motivate behaviour change across diverse domains. However, no empirical evidence
exists on whether or not persuasive strategies lead to certain emotions to inform which
strategies are most appropriate for delivering interventions that not only motivate users
to perform target behaviour but also help to regulate their current emotional states. We
conducted a large-scale study of 660 participants to investigate if and how individ-
uals including those at different stages of change respond emotionally to persuasive
strategies and why. Specifically,we examined the relationship between perceived effec-
tiveness of individual strategies operationalized in a system and perceived emotional
states for participants at different stages of behaviour change. Our findings estab-
lished relations between perceived effectiveness of strategies and emotions elicited in
individuals at distinct stages of change and that the perceived emotions vary across
stages of change for different reasons. For example, the reward strategy is associated
with positive emotion only (i.e. happiness) for individuals across distinct stages of
change because it induces feelings of personal accomplishment, provides incentives
that increase the urge to achieve more goals, and offers gamified experience. Other
strategies are associated with mixed emotions. Our work links emotion theory with
behaviour change theories and stages of change theory to develop practical guidelines
for designing personalized and emotion-adaptive persuasive systems.
BOladapo Oyebode
oladapo.oyebode@dal.ca
Darren Steeves
darren@jackhabbit.com
Rita Orji
rita.orji@dal.ca
1Faculty of Computer Science, Dalhousie University, Halifax, NS B3H 1W5, Canada
2School of Health and Human Performance, Dalhousie University, Halifax, NS B3H 4R2, Canada
123
O. Oyebode et al.
Keywords Persuasive strategies ·Emotional states ·Emotion-adaptive persuasive
systems ·Adaptive interventions ·Behaviour change ·Stages of change ·
Adaptivity ·Personalization ·Tailoring ·Mental health ·Health and well-being ·
Design guidelines
1 Introduction
Persuasive strategies have been operationalized in applications to achieve behaviour
change objectives across diverse domains. In the health domain, for example, human—
computer interaction (HCI) researchers have designed persuasive systems using
various persuasive strategies to promote physical activity (Ciman et al. 2016; Haque
et al. 2020; Lee et al. 2018; Oyebode et al. 2020; Saksono et al. 2020) and healthy
eating (Hsu et al. 2014;Orjietal.2013b), discourage substance use (Canale et al.
2015; Klein et al. 2019; Puddephatt et al. 2019; Schulz et al. 2013) and risky sexual
behaviour (Ndulue and Orji 2018), as well as manage diseases (Rezai and Burns 2014;
Schnall et al. 2015) and combat mental health issues (Arshad et al. 2019; Bardram
et al. 2012; Fuller-Tyszkiewicz et al. 2018; Khan and Peña, 2020). Previous research
suggests that technologies can create outlets for expressing emotions (Shank 2014)
and that psychophysiological assessments could be useful for personalizing persuasive
technologies (Spelt et al. 2022); yet, there is no empirical evidence in the literature
regarding whether or not persuasive strategies operationalized in persuasive systems
elicit emotions in users across different stages of behaviour change. Emotional states
have been shown to influence physical and mental health (Gross and Muñoz 1995;
Salovey et al. 2000). For example, negative emotional states create unhealthy pat-
terns of physiological functioning (Salovey et al. 2000) and inhibit resilience and
quality of life (Philippe et al. 2009). In addition, people exhibiting negative emo-
tions are at increased risk of illness, immunity relapse, and mortality (Natt och Dag
et al. 2020; Salovey et al. 2000). Conversely, positive emotions enhance physical
health (e.g. cardiovascular health (Boehm et al. 2020)), social interaction, cognition,
and psychological well-being including resilience and sense of identity (Alexander
et al. 2021; Ching and Chan 2020; Fredrickson and Joiner 2002; Gloria and Steinhardt
2016). This highlights the need for delivering interventions that promote positive emo-
tions, in line with studies which show that inability to adequately address challenging
emotions is associated with mental health issues including depression, substance use,
anxiety and mood disorders, eating disorders, and personality disorder (Berking and
Wupperman 2012). To achieve this, it is imperative to investigate if there are rela-
tions between perceived effectiveness of persuasive strategies and emotions elicited
in individuals to inform design decisions that support emotion-adaptive interventions
in persuasive systems; however, this is yet to be investigated via an empirical study.
As mentioned previously, persuasive strategies motivate behaviour change; hence,
people’s current stage of change should also be considered to effectively tailor emotion-
adaptive behaviour change interventions since individuals progress through multiple
stages of change when trying to modify their behaviours (Prochaska 2020).
Therefore, we conducted an empirical study to investigate whether or not the per-
ceived effectiveness of persuasive strategies operationalized in a system leads to
123
Persuasive strategies and emotional states: towards designing …
positive and/or negative emotions across different stages of behaviour change and
why. To achieve this, we examined the relationship between perceived effectiveness
of individual strategies operationalized in a system and perceived emotional states
for participants at different stages of behaviour change. We used four well-established
theoretical frameworks to achieve our research goal: Ekman’s emotion theory (Ekman
1992,1999; Ekman and Ekman 2016), persuasive systems design (PSD) model (Oinas-
Kukkonen & Harjumaa 2009), App Behaviour Change Scale (ABACUS) (McKay et al.
2019), and the transtheoretical model of behaviour change (TTM) (Prochaska et al.
1993). The Ekman’s emotion theory consists of six basic emotional states out which
five are universal (i.e. “emotions that all humans, no matter where or how they are
raised, have in common”), according to Ekman’s atlas of emotion (Ekman and Ekman
2016). The five universal emotional states, which are used in this work, comprise four
negative emotional states (anger, fear, disgust, and sadness) and a positive emotional
state (happiness). Ekman’s emotion model has been very popular in the literature over
the years, including in HCI research (Elgarf mahaeg et al. 2021; Krekhov et al. 2022;
Zhang et al. 2018). The TTM framework posits that people progress through six stages
to adopt healthy behaviour and that interventions need to be tailored to support the
motivational needs of individuals at each stage (Prochaska et al. 1993,2015; Prochaska
and Velicer 1997). TTM has been widely applied in several HCI researches includ-
ing(Leeetal.2017; Lin et al. 2019; Mulchandani et al. 2022; Oyebode et al. 2021;
Oyebode and Orji 2022). These stages of behaviour change include Precontemplation,
Contemplation, Preparation, Action, Maintenance, and Termination. Furthermore, the
PSD model is popularly known for its 28 persuasive strategies for designing and eval-
uating persuasive or behaviour change systems and has enjoyed widespread use in
HCI especially persuasive technology research (Ganesh et al. 2021; Oduor and Oinas-
Kukkonen 2019;Orjietal.2018b; Oyebode et al. 2021; Oyebode and Orji 2022).
Similarly, the ABACUS framework consists of 21 persuasive strategies for assessing
the behaviour change potential of smartphone applications or systems (McKay et al.
2019).
To explore how to design emotion-adaptive persuasive systems and effectively tai-
lor them based on users’ current stage of change, we conducted a large-scale study of
660 participants to investigate if and how individuals at different stages of behaviour
change respond emotionally to eleven (11) persuasive strategies and why. The strate-
gies include self-monitoring,reminder,reduction,rehearsal,goal setting,suggestion,
reward,expertise,opportunities to plan for barriers,distraction or avoidance, and
recognition. Our choice of strategies was inspired by an earlier study which combined
the PSD and ABACUS strategies and then ranked them in terms of their behaviour
change score—BCS (a measure of the extent to which a strategy is implemented
in apps) (Alslaity et al. 2022). For this work, we selected strategies that have high
BCS and also those with low BCS since rarely used strategies could also be effective
(Alslaity et al. 2022). To investigate how individuals across the six stages of change
respond emotionally to the strategies, we examined the relations between perceived
effectiveness of each strategy and Ekman’s universal emotional states (anger,fear,
disgust,happiness, and sadness) for people at different stages of change. We used the
structural equation modelling technique (Hair et al. 2011) to create models showing
123
O. Oyebode et al.
the relationship between perceived effectiveness of the strategies and perceived emo-
tional states, and then conducted multi-group analysis to explore for differences across
the models for people at distinct stages of change. To collect data from participants,
we iteratively created and presented eleven low-fidelity prototypes illustrating each
strategy within the context of resilience building (which refers to the mental ability
to bounce back or cope with adversities such as everyday stressors and traumatic life
events (VanMeter and Cicchetti 2020)), followed by validated scales measuring per-
ceived effectiveness (Thomas et al. 2019) of each strategy and perceived emotions per
strategy. We also conducted thematic analysis of participants’ qualitative comments
to understand why each persuasive strategy leads to specific emotion(s) in participants
across the various stages of change.
Our findings established relations between perceived effectiveness of persuasive
strategies and emotions elicited in individuals at distinct stages of change and that the
perceived emotions vary across stages of change for different reasons. For example,
the Reward strategy is significantly and strongly associated with happiness emotional
state across all stages of change because it induces feelings of personal accomplish-
ment, provides incentives that increase the urge to achieve more goals, and offers
gamified experience. Also, Reminder strategy is significantly associated with fear for
participants at the Preparation stage of change because they were afraid of receiving
multiple or excessive reminders that could induce anxiety or become overwhelming.
Our findings could inform designers about which persuasive strategies to employ and
those to avoid when designing emotion-adaptive persuasive systems targeting users at
different stages of behaviour change.
This paper offers four contributions in the fields of emotion-driven adaptation
and persuasive technology design in HCI. First, we underpin the significance of
people’s emotional responses to persuasive strategies and establish that emotional
states are important dimensions for tailoring and selecting appropriate strategies
to improve the effectiveness of persuasive systems. Second, we provide empirical
insights into the relationship between perceived effectiveness of persuasive strate-
gies and perceived emotional states across different stages of behaviour change using
four well-established theoretical frameworks. Third, we offer qualitative insights to
explain why individual strategies lead to specific emotion(s) across different stages
of change based on participants’ comments. Fourth, we reflect on our findings and
offer practical guidelines for designing persuasive systems that are emotion-adaptive
such that behaviour change interventions are delivered using appropriate persuasive
strategies that promote positive emotion (and reduce negative emotions) based on
users’ current stage of change. To the best of our knowledge, our work is the first to
link emotion theory (from Ekman’s emotion model) with behaviour change theories
(from PSD and ABACUS frameworks) and stages of change theory (from TTM) to
find patterns in people’s emotional responses to persuasive strategies at distinct stages
of change that can inform the choice of strategies to employ in designing and tailoring
emotion-adaptive persuasive systems to motivate behaviour change.
123
Persuasive strategies and emotional states: towards designing …
2 Related work
This section presents a literature review of emotion theories, persuasion and persua-
sive strategies, the transtheoretical model and its application in HCI, emotion-based
persuasion, and emotion-based adaptation in HCI.
2.1 Emotion theories
Emotion theories define how emotions are represented in applications or systems.
These theories are broadly classified into discrete and dimensional categories. The
discrete emotion theories place emotion into unique categories or states. Discrete
theories of emotion (such as Ekman’s emotion theory) are the most adopted in the
literature for emotion detection tasks (Nandwani and Verma 2021). Ekman’s emotion
theory (Ekman 1992,1999) posits that there exist six basic and distinct emotional
states: happiness, sadness, anger, fear, disgust, and surprise. The six basic emotions
stem from Ekman’s argument that emotions evolved for their adaptive value in dealing
with fundamental life tasks which are universal human predicaments or common
adaptational tasks like achievements, loses, frustrations, immediate danger, and so on
(Ekman 1999). Most researchers, however, did not consider surprise an emotion since
there is no clear evidence that it is positive or negative (Ekman 1992; Tracy and Randles
2011). In addition, surprise is tagged as a controversial emotion (Tracy and Randles
2011), excluded from established positive emotions (Fredrickson 1998; Sauter 2010)
and even shown to have a negative valence (Noordewier and Breugelmans 2013). In
the Ekman’s atlas of emotions, happiness/enjoyment, sadness, anger, fear, and disgust
were highlighted as the “five universal emotions: emotions that all humans, no matter
where or how they are raised, have in common” (Ekman and Ekman 2016).
Plutchik emotion theory (Plutchik 1980,1982) extends Ekman’s basic emotions by
including two additional emotional states: trust and anticipation. The Orthony, Clore,
and Collins (OCC) emotion theory (Ortony et al. 1990) discretized emotion into 22
categories that also include Ekman’s basic emotions. Parrot emotion theory (Parrott
2001) considers Ekman’s basic emotions as primary emotions and then extends them
to 100 secondary emotions arranged in a tree structure. On the other hand, dimensional
emotion theories position emotions on a dimensional space showing how emotions are
related, as well as their intensities or degree of occurrence. A popular example is the
Russell’s Circumplex Model of Affect which projects emotions in two dimensions:
valence (pleasure-displeasure continuum) and arousal, with each emotion posited as
a linear combination (or as varying degrees) of both dimensions (Posner et al. 2005;
Russell 1980). A third dimension—dominance (which describes the degree to which
individuals had control over their emotion) was added by Russell and Mehrabian
(Russell and Mehrabian 1977).
Of the emotion theories, the discrete theories have enjoyed widespread use in HCI
research (Elgarf mahaeg et al. 2021; Krekhov et al. 2022; Wagener et al. 2022; Zhang
et al. 2018), compared to the dimensional theories. We focus on Ekman’s five universal
emotions (Ekman and Ekman 2016)—happiness,sadness,anger,fear , and disgust—in
this work.
123
O. Oyebode et al.
2.2 Persuasion and persuasive strategies
Persuasive or behaviour change systems employ various persuasive strategies to moti-
vate behaviour change and have been shown to be effective (Hamari et al. 2014;
Orji and Moffatt 2018). Human–computer interaction (HCI) researchers targeting
various domains, including health, have increasingly operationalized these strategies
in their systems or designs to motivate users and achieve desirable behaviours. For
instance, persuasive systems for health, which are designed with the intent of promot-
ing healthy behaviours or discouraging risky behaviours (Fogg and Fogg 2003;Orji
et al. 2014,2017), have helped people improve personal wellness (such as physical
activity, healthy eating, and weight management) (Arteaga et al. 2009; Curtis et al.
2015; Haque et al. 2020; Hsu et al. 2014; Lieffers et al. 2018; Purpura et al. 2011),
overcome addictive and risky behaviours (e.g. substance abuse) (Canale et al. 2015;
Klein et al. 2019; Ndulue and Orji 2018; Puddephatt et al. 2019; Schulz et al. 2013), as
well as manage diseases (Rezai and Burns 2014; Schnall et al. 2015) and mental health
issues (Arshad et al. 2019; Bardram et al. 2012; Fuller-Tyszkiewicz et al. 2018; Khan
and Peña, 2020). Over the years, persuasive technology researchers have proposed
several persuasion frameworks offering persuasive strategies that could bring about
behaviour change. Popular frameworks include Persuasive Technology Tools (Fogg
and Fogg 2003), Cialdini’s Principles of Persuasion (Cialdini 2001), and persuasive
systems design (PSD) framework (Oinas-Kukkonen and Harjumaa 2009). The PSD
framework, which comprises 28 persuasive strategies, is grounded in psychological
and behavioural theories, extends the other two frameworks, and more importantly is
aimed at facilitating the development and evaluation of persuasive systems. Another
persuasion framework is the App Behaviour Change Scale (ABACUS) consisting of
21 strategies for assessing the behaviour change potential of smartphone applications
or systems (McKay et al. 2019). A study by some HCI researchers combined the PSD
and ABACUS strategies to assess the persuasiveness of mental health applications and
then ranked the persuasive strategies in terms of their behaviour change score—BCS
(a measure of the extent to which a strategy is implemented in the apps) (Alslaity
et al. 2022). This study inspired our choice of eleven (11) strategies for this work (see
Table 1) by considering not only strategies with high BCS but also those with low
BCS since rarely used strategies could also be effective (Alslaity et al. 2022).
2.3 The transtheoretical model and HCI
The transtheoretical model of change (TTM) is a popular theory for modelling
behaviour change as a dynamic process involving various stages through which indi-
viduals progress until they achieve their desired behaviour (Prochaska et al. 1993).
The TTM builds upon diverse theories of psychotherapy and has been shown to be one
of the most widely employed theories of health behaviour change (Evers et al. 2006,
2012; Ferron and Massa 2013;Horwath1999; Johnson et al. 2008; Lee et al. 2017).
The TTM consists of six stages of behaviour change that individuals progress through:
Precontemplation,Contemplation,Preparation,Action,Maintenance, and Termina-
tion. People at the Termination stage have adopted the healthy behaviour as part of
123
Persuasive strategies and emotional states: towards designing …
Table 1 Description of persuasive strategies and the corresponding theoretical framework(s)
Strategy Framework(s) Description
Self-monitoring PSD and ABACUS Provides means for users to track their own
progress or performance towards their target
behaviour or goal
Reminder PSD and ABACUS Reminds or notify users of their target behaviour
during system use
Reduction PSD Reduces efforts required to perform the target
behaviour
Rehearsal PSD and ABACUS Provides means of rehearsing or practicing the
target behaviour
Goal setting ABACUS Allows users to set a goal that can be achieved
during system use
Suggestion PSD Provides tailored suggestions or tips for
achieving the desired behaviour during system
use
Reward PSD and ABACUS Incentivizes users for achieving specific
milestones using virtual rewards such as
badges, points, etc.
Expertise PSD and ABACUS Provides content showing knowledge,
experience, and competence
Opportunities to plan for
barriers
ABACUS Encourages users to think about potential
barriers and identify ways of overcoming them
Distraction or Avoidance ABACUS Gives suggestions and advice on how users can
avoid situations or distract themselves when
trying to reach their goal
Recognition PSD Provides public recognition for users who
perform the target behaviour
their daily lifestyle with no intention to relapse (Prochaska et al. 1993). To address the
“one-size-fits-all” approach profound in most commercial behaviour change technolo-
gies (He et al. 2010) where all users receive the same intervention, HCI researchers
have applied the TTM to design and evaluate behaviour change interventions that con-
sider users’ current state of change as part of the design/evaluation process (Ferron &
Massa 2013;Horwath1999; Johnson et al. 2008). Table 2describes the six (6) TTM
stages of change within the domain of resilience building which is our use case in this
paper.
2.4 Influencing persuasion using emotion
Emotion or affect has been shown to influence persuasion and that different emo-
tions produce different persuasion outcomes (Dillard and Seo 2012; Hamby and Jones
2022). Majority of the research in this area induce certain emotions into messages
or communications to enhance their persuasive power. Specifically, loss-framed mes-
sages induce negative emotions (e.g. fear) while gain-framed messages induce positive
123
O. Oyebode et al.
Table 2 Transtheoretical model (TTM) stages of behaviour change
Stage of Change Description
Precontemplation People at this stage have no intention to become resilient in the future
Contemplation Contemplators have not taken steps to become resilient but are planning to start
within the next 6 months
Preparation People at this stage are seriously considering to become resilient and have taken
some steps towards achieving this objective
Action Individuals at this stage have actively been working towards becoming resilient
for up to 6 months
Maintenance People at this stage have become resilient in their daily lives for more than
6 months
Termination Being resilient is already the habit of people at this stage. They have been resilient
in their daily lives for several years
emotions (e.g. humour,hope,happiness, empathy) to convince target audience to
perform specific behaviours (van ’T Riet et al. 2010). For example, loss-framed or
fear-induced insurance advertisements (ads) could depict the threats or dangers posed
by lack of coverage (of lives and properties) from disasters to change people’s per-
spectives and ultimately boost sales; however, mixing fear and humour would reduce
defensive responses to these ads and increase persuasion (Mukherjee and Dubé 2012).
Similarly, mixed emotions (i.e. combining positive and negative emotions) in ad mes-
sages discouraging drink driving were found to be more effective in changing drink
driving attitudes, norms, and intentions of young adults (Yousef et al. 2021). This
aligns with research evidence on the efficacy of mixed emotions for controlling risky
behaviours (Carrera et al. 2010). Furthermore, Appel et al.’s emotional storytelling
robot, which facially displays the emotion congruent to the semantics of the story-
based ad, significantly and positively influenced individuals’ decision to purchase the
product advertised in the story (Appel et al. 2021).
In the area of health and wellness, loss-framed messages were shown to increase
people’s intention to engage in healthy behaviour such as reducing salt intake (van
’T Riet et al. 2010) or getting vaccinated against diseases (Ye et al. 2021), com-
pared to gain-framed messages. In contrast, gain-framed or positively valenced health
messages promoting healthy eating and physical activity were found to evoke pos-
itive attitude and improve self-efficacy (Ort et al. 2021). Cho et al. further revealed
that fear induction in public health campaigns (such as in skin cancer prevention
messages) could lead to stronger or weaker intentions to engage in recommended
behaviour for individuals at different stages of change (Cho and Salmon 2006). In the
area of sustainable environment, Ibanez et al.’s experiments revealed that inducing
hope into climate change efficacy messages boost supportive attitudes and advocacy
behaviour (Nabi et al. 2018). A similar study found negative emotions (such as sad-
ness and shame) as significant barriers to individuals’ willingness to render monetary
donations to environmental non-governmental organizations (ENGOs), compared to
positive emotions (happiness and pride) (Ibanez and Roussel 2021). In line with this,
123
Persuasive strategies and emotional states: towards designing …
Samad et al. induced empathy in their dialog system to persuade users to give dona-
tions (Samad et al. 2022). Rather than focusing on positive emotions, Homer argued
that combining hope and sadness would elicit sympathy and inspiration that motivate
increased donations towards human-suffering causes (Homer 2021).
2.5 Emotion-based adaptation in HCI
Adaptation based on emotional state is receiving increasing attention in HCI research
and across domains including health, education, entertainment, and transportation.
For instance, Ghandeharioun et al. aimed to improve emotional well-being of peo-
ple through their emotion-aware chatbot (Ghandeharioun et al. 2019). The agent
periodically prompts users to rate their mood, and then randomly draws emotion-
ally relevant phrases scripted for the reported state to provide adaptive empathetic
response. Similarly, the emotion-adaptive chatbot (GremoBot) by Peng et al. (2019)
reinforces positive feelings in team-based collaboration while steering group members
away from negative words. GremoBot analyses text messages for negative emotions
and then provide tips to help members interpret the issue in a positive way. To improve
medication adherence in hypertensive patients, Condori-Fernandez et al. developed
an adaptive mobile application by delivering appropriate persuasive messages to
patients whenever their current emotional state is stressed based on physiological
data (Condori-Fernandez & Lopez 2017). In the area of entertainment, Ibanez et al.
developed an audio system for video games to improve player engagement by adapting
music to game situation based on current emotional state (i.e. anger,disgust,fear,
happiness,sadness,orsurprise) (López Ibáñez et al. 2018). The system combines
short video tracks in real-time using current emotion as cue. Similarly, Frommel et al.
developed a framework for adapting game features (such as content and difficulty
level) based on players’ current emotional state (Frommel et al. 2018a;b) to improve
their gaming experience (i.e. making it more enjoyable and engaging).
In automotive environments, Braun et al. compared various emotion-adaptive auto-
motive interfaces (ambient light, visual notification, voice assistant, and empathic
voice-enabled assistant) to regulate drivers’ emotions while driving (Braun et al. 2019).
The empathic assistant, which provides empathetic messages tailored to angry or sad
drivers, was found to be most effective. A similar system, named “adaptive voice alert
system”, changes voice alerts depending on the current mood of the driver (Sarala
et al. 2018). For example, the voice alert is intense when the driver is sad,angry,dis-
gusted,orafraid, but moderate if current emotion is positive (e.g. happiness). To foster
learning, HCI researchers developed systems (e.g. for practicing coding) which pro-
vides personalized and adaptive content (such as controlling the complexity of practice
questions/problems and offering problem-specific suggestions or tips) whenever a stu-
dent is in the confused state (Tiam-Lee and Sumi 2018) or exhibiting one of Ekman’s
emotions or contempt (Taurah et al. 2020). Similarly, Rodriguez et al. developed an
e-learning system with the aim of increasing student performance/engagement dynam-
ically (Rodriguez et al. 2014). The system detects individual students’ emotional state
(joy,anger,sadness,orfear) from their self-written essays and then recommend
activities and adapt content based on the emotional state.
123
O. Oyebode et al.
2.6 Research goal
Persuasive strategies have been widely operationalized in applications or systems to
motivate behaviour change across diverse domains without coercion or deception.
Although emotion has been used to influence persuasive outcomes (see Sect. 2.4),
they are usually coercive or manipulative (such as inducing fear in advertising mes-
sages to drive product sales) which is against persuasion principles according to Fogg
and Fogg (2003). In addition, such persuasion mechanisms are mostly limited to mes-
sages/communications and may fail to influence behaviour change in non-messaging
contexts. Moreover, emotion-adaptive systems provide interventions that regulate
users’ current emotion to achieve desired outcomes (see Sect. 2.5); hence, opera-
tionalizing persuasive strategies in these systems will increase the persuasive power
of interventions in motivating users to adopt the target behaviour while regulating
their current emotional state. This means that persuasive strategies must align with
(or support) the emotion regulation goal of such systems—which is to promote posi-
tive emotions and reduce negative emotions. However, no empirical evidence exists on
how people respond emotionally to persuasive strategies operationalized in persuasive
systems to inform which strategies are most appropriate for delivering interventions
that not only motivate users to perform target behaviour but also help to regulate
their emotional state. To address this gap, we conduct a large-scale empirical study to
investigate whether or not the perceived effectiveness of strategies operationalized in a
system leads to positive and/or negative emotions across different stages of behaviour
change and why. Considering that there are many persuasive strategies and emotional
states in the literature, modelling which emotions are elicited by which strategies for
each stage of change could be challenging; yet, our empirical study established a
relationship between perceived effectiveness of persuasive strategies and perceived
emotions elicited in individuals at distinct stages of behaviour change which is an
important first step towards a broader research in this area. Our work links emotion
theory with behaviour change theories and stages of change theory to develop practical
guidelines for designing emotion-adaptive persuasive systems that employ strategies
that are most suitable for delivering behaviour change interventions for users across
different stages of change while promoting positive emotions.
3 Methodology
To achieve our research goal, we followed the methodological procedure below:
1. We selected eleven (11) persuasive strategies from the PSD and ABACUS persua-
sion frameworks (see Table 1). The strategies included not only those with high
behaviour change scores but also those with low scores (see Alslaity et al. 2022)
since rarely used strategies could also be effective (Alslaity et al. 2022).
2. We designed low-fidelity prototype illustrating each persuasive strategy. We con-
textualized the prototypes in the mental health domain of resilience building which
encompasses various mechanisms for coping with adverse life experiences such
as trauma and stress (Pfefferbaum et al. 2008; Shastri 2013).
123
Persuasive strategies and emotional states: towards designing …
3. We conducted an online study to assess the perceived persuasiveness/effectiveness
of the strategies and then identified perceived emotions per strategy for participants
at different stages of change.
4. Next, we analysed the quantitative and qualitative data collected using well-
established analytical methods or techniques including structural equation mod-
elling and thematic analysis.
5. Finally, based on our findings, we offer practical guidelines for designing emotion-
adaptive systems that employ strategies that are most appropriate for delivering
behaviour change interventions for users at different stages of change while rein-
forcing positive emotion.
We discuss the details of our methodology including study design and data analysis
in subsequent subsections.
3.1 Study design
We followed established methodologies employed in several HCI researches to collect
data for our study (Jia et al. 2016;Orjietal.2017,2018b; Oyebode et al. 2021), as
described below.
3.1.1 Prototype development
To assess the perceived persuasiveness/effectiveness of the strategies (i.e. self-
monitoring,reminder,reduction,rehearsal,goal setting,suggestion,reward ,exper-
tise,opportunities to plan for barriers,distraction or avoidance, and recognition), we
created a low-fidelity prototype that illustrates each strategy as a persuasive feature in a
mobile app for building human resilience. Our decision to contextualize the prototypes
in the domain of resilience building is based on an evidence-based connection between
resilience (which is the mental ability to bounce back or cope with adversities such
as everyday stressors and traumatic life events (VanMeter and Cicchetti 2020)) and
emotion regulation (Polizzi and Lynn 2021). Specifically, promoting positive emo-
tions (such as happiness) and reducing negative emotions (such as sadness, anger,
fear, etc.) have been shown to improve resilience (Cohn et al. 2009; Richardson et al.
2016; Short et al. 2018; Tugade and Fredrickson 2007). We collaborated with ten (10)
experts in psychology, resilience training, and persuasive technology to identify var-
ious resilience building mechanisms that can be delivered using the strategies. These
prototypes (one per strategy) were evaluated twice by the experts, and their feedback
were used in refining the prototypes. We designed the prototypes such that they are
easily understood by target audience from diverse backgrounds. Prototype design is
a common practice in HCI research to showcase implementation ideas for evaluation
purposes (Jalowski 2020; Jia et al. 2016; Oppenlaender et al. 2021; Oyebode et al.
2021; Oyebode and Orji 2022). We created our prototypes using the Balsamiq app
prototyping tool (Balsamiq, n.d.). Figures 1and 2show the prototypes illustrating
Reward and Reduction strategies, respectively.
123
O. Oyebode et al.
Fig. 1 Prototype illustrating the
Reward strategy
3.1.2 Experimental design
The online study, which is a within-subject study, is depicted in Fig. 3. First, we asked
participants to complete demographic questions including age range, gender, marital
status, level of education, and profession. They were also asked to complete a TTM-
based questionnaire (adapted from Gonzalez-Ramirez et al. 2017), which is composed
of six (6) multiple choice questions, to indicate their current stage of behaviour change
in the context of resilience building (see Table 2). Next, we asked participants to
watch a short neutral video prior to examining each prototype to remove bias (or
carryover effect) from the emotional states reported. This approach has been used in
HCI affective experiments including (Soleymani et al. 2012). The short neutral video
clips used in this study were randomly selected from the film library for affective
scientists provided by Stanford Psychophysiology Laboratory (Samson et al. 2016).
Next, we presented a prototype illustrating each strategy to participants as a set of
123
Persuasive strategies and emotional states: towards designing …
Fig. 2 Prototype illustrating the
Reduction strategy
images in a logical flow depicting user interaction. Participants see one prototype at a
time and then respond to questions that assess the perceived effectiveness (or perceived
persuasiveness) of the prototype/strategy. To prevent possible prototypes ordering
bias, we used the randomization feature of the survey tool to vary the position (i.e.
change the order) of the prototypes for each participant. The perceived persuasiveness
questionnaire (PPQ) (Thomas et al. 2019), which was adapted to fit our resilience
building context, has been used in many HCI and persuasive technology research
including (Orji et al. 2014,2017,2018b; Oyebode et al. 2021; Oyebode and Orji
2022; Wais-Zechmann et al. 2018). The PPQ consists of five (5) items measured using
a 7-point Likert scale ranging from “1—Strongly Disagree” to “7—Strongly Agree”:
(a) This application would influence me to be resilient,(b)This application would
convince me to be resilient,(c)This application would be personally relevant for me,
123
O. Oyebode et al.
Fig. 3 Stages involved in the online study
(d) This application would make me to reconsider my resilience level, and (e) The
strategy would make or motivate me to use the application.
Next, participants were asked to indicate how they felt (i.e. their emotional state)
while imagining themselves using the application depicted in the prototype. Partic-
ipants rated the degree to which they experienced each of Ekman’s five universal
emotions (Ekman and Ekman 2016)—happiness,sadness,anger,fear , and dis-
gust—on a 5-point Likert scale ranging from “1—Strongly Disagree” to “5—Strongly
Agree”, an approach used in emotion-based HCI research including (Wilberz et al.
2020) in line with research evidence that supports the use of single-item scales
for assessing discrete emotions (Abdel-Khalek 2006; Gross and Levenson 1993).
Afterwards, we asked participants to justify their ratings by providing qualitative
comments. We included attention-check questions to ensure that participants were
actively considering their responses. The online study was designed using the Opinio
tool (ObjectPlanet Inc., n.d.). Prior to the actual study, we conducted two pilot studies.
The first pilot study was conducted with 20 random students recruited from a univer-
sity and the second pilot study with 40 participants recruited from MTurk to test the
validity of our study instruments.
123
Persuasive strategies and emotional states: towards designing …
3.1.3 Participants
We recruited participants for our large-scale online study using multiple channels
including email, social media, snowball sampling, SONA, and Amazon Mechanical
Turk (MTurk). SONA is a widely used experimental participation system that allows
university students to participate in research studies and earn credit points that can
be added to their final grade (Sona Systems Ltd, n.d.). MTurk is an established and
reliable method of recruiting large and diverse participants and has been used in many
HCI studies including (Hasan et al. 2021; Jia et al. 2016; Koshy and Park 2021;Orji
et al. 2018b; Oyebode et al. 2021). In addition, the MTurk platform efficiently and
securely distributes surveys to global audience at a reasonable cost with high quality
(Buhrmester et al. 2015; Mason and Suri 2012). Having removed incomplete survey
responses, as well as responses containing inaccurate answers to comprehension and
attention-based questions (Mason and Suri 2012), we included a total of 660 valid
responses in our analysis. In line with our research ethics approval, eligibility or
inclusion criteria require that participants be adults (18 years or older) and proficient
in English language. As shown in Table 3, our participants were diverse in terms of
age, gender, marital status, education, profession, and their current stage of behaviour
change.
3.2 Data analysis
We used well-established analytical techniques and tools to analyse the quantitative
and qualitative data collected, as summarized below:
1. We used the Kaiser–Meyer–Olkin (KMO) Measure of Sampling Adequacy and the
Bartlett Test of Sphericity (BTS) methods (Kaiser 1970) to determine whether our
Table 3 Demographic information of participants
Total participants 660
Age 18–25 (42%), 26–35 (30%), 36–45 (17%), Over 45 (11%)
Gender Male (39%), Female (60%), Other (1%)
Marital Status Single (42%), Married (52%), Widowed (0%), Divorced (2%), Separated (1%),
Registered Partnership (1%), Other (2%)
Education High school or equivalent (7%), College diploma (3%), Bachelor’s degree (52%),
Master’s degree (34%), Doctoral degree (1%), Other (1%)
Profession (Top 15) College, University, and Adult Education (22%); Finance and Insurance (10%);
Others (10%); Information Services and Data Processing (10%);
Manufacturing (9%); Health Care and Social Assistance (8%); Computer and
Electronics Engineering (8%); Software (4%); Arts, Entertainment, and
Recreation (4%); Retail (3%); Primary/Secondary Education (2%);
Construction (2%); Hotel and Food Services (2%); Agriculture, Forestry,
Fishing and Hunting (2%); Scientific or Technical Services (2%)
Stage of Change Precontemplation (8%), Contemplation (9%), Preparation (37%), Action (20%),
Maintenance (9%), and Termination (18%)
123
O. Oyebode et al.
data are suitable for factor analysis. KMO measures the sampling adequacy of the
variables, while BTS determines if certain redundancy exists between the variables
that can be summarized with fewer factors. A good factor-analytic data should
have a minimum KMO value of 0.8, while the recommended minimum value is
0.6 (Kaiser 1970). These tests are essential prior to multivariable modelling.
2. Next, we used the partial least square structural equation modelling (PLS-SEM)
method to model the relationship between the strategies and the five emotional
states. We chose PLS-SEM over other methods (e.g. covariant-based) because it
is highly suitable for creating complex models (Kupek 2006) and has been widely
used in estimating relationships between latent and observed variables in many
HCI studies including (Jicol et al. 2021; Machuletz et al. 2018;Orjietal.2018b;
Oyebode et al. 2021; Oyebode and Orji 2022; Wang et al. 2018). The structural
model in Fig. 4shows the relations between the Reward strategy and the five
emotional states. Latent variables include each strategy (e.g. Reward strategy) and
the five emotional states—happiness,sadness,anger,fear, and disgust, while
observed variables are the five items or indicators measuring perceived persua-
siveness of the strategy and the item measuring each emotional state. We used the
SmartPLS tool (Sarstedt and Cheah 2019; SmartPLS GmbH, n.d.) to create the
models, one strategy at a time.
3. We also performed multi-group comparison followed by Bonferroni adjustment
to explore for differences across the models for people at distinct stages of change
and opportunities for mapping perceived effectiveness of strategies to perceived
evoked emotions based on people’s stage of change using the multi-group analysis
functionality and procedure in SmartPLS 3 (Matthews 2017; Sarstedt et al. 2011).
Before comparing our models, we established measurement invariance across the
six stages of change and also performed model validity and reliability checks (Hair
et al. 2017b; Henseler et al. 2016), an approach already used in HCI research (Orji
et al. 2013a; Oyebode et al. 2021). We provide further details on each of these
analysis in Sects. 3.3 and 3.4.
4. Finally, we applied thematic analysis method to analyse participants’ qualitative
comments to support their quantitative ratings. This is aimed at extracting qual-
itative insights that explain why individual strategies lead to specific emotion(s)
across different stages of behaviour change.
3.3 Measurement validation
As highlighted in Sect. 3.2, we determined the appropriateness of our data for fac-
tor analysis using the KMO Measure of Sampling Adequacy and the Bartlett Test of
Sphericity (BTS). Our results revealed that the KMO value was 0.968, well above
the recommended value of 0.6. In addition, the BTS was statistically significant:
χ2(11175) 110547.633, p< 0.0001. These results show that our dataset is suited
for factor analysis.
Prior to multi-group comparison, we need to establish that we are not comparing
dissimilar groups. Hence, we established measurement invariance following a three-
step procedure established for PLS-SEM (Henseler et al. 2016). First, we established
123
Persuasive strategies and emotional states: towards designing …
Fig. 4 Structural model depicting the relationship between the Reward strategy and the five emotional states
configural invariance which ensures that the same basic factor structure exists in all the
groups. Second, we assessed compositional invariance (i.e. equal indicator weights),
and finally, we established the equality of composite mean values and variances across
groups. We used the component-based Confirmatory Factor Analysis (CFA) feature
in SmartPLS 3 to conduct factor analysis for each group of data and retained items
that had factor loadings of at least 0.5 (Hair et al. 2011) in all the groups, thereby
establishing configural invariance. We also establish compositional invariance and
equality of composite mean values and variances across groups following the PLS-
SEM procedure in Henseler et al. (2016).
Next, we validated the measurement models using the criteria suggested by Chin
(1998) before estimating the structural paths to test for the relationship between the
variables. Specifically, we assessed the property of scales in terms of: (1) indicator
reliability (using Cronbach’s alpha and composite reliability), (2) convergent reliability
(using the Average Variance Extracted (AVE) metric), and (3) discriminant validity
based on the heterotrait–monotrait (HTMT) ratio of correlations. For all models, the
indicator reliability can be assumed since both Cronbach’s alpha (for the multi-item
construct) and composite reliability that assess internal consistency and the strength of
each indicator’s correlation with its latent variables are higher than the recommended
value of 0.7 (Chin 1998;Hairetal.2017a). Also, the convergent reliability can be
assumed since the AVE values which represent the variance extracted by the variables
from their indicator items were above the recommended value of 0.5 (Chin 1998;Hair
et al. 2017a). Finally, the discriminant validity can be assumed since the HTMT values
for accessing discriminant validity were below the recommended limit of 0.9 (Hair
123
O. Oyebode et al.
Table 4 Cronbach’s alpha, composite reliability, and AVE values of the constructs
Construct Cronbach’s alpha
(≥0.7)
Composite reliability
(≥0.7)
AVE
(≥0.5)
Anger – 1.000 1.000
Disgust – 1.000 1.000
Fear – 1.000 1.000
Happiness – 1.000 1.000
Sadness – 1.000 1.000
Reward strategy (Persuasiveness) 0.932 0.949 0.787
Table 5 HTMT values of the five emotional states and the Reward strategy’s persuasiveness
Construct Anger Disgust Fear Happiness Reward strategy
(Persuasiveness)
Sadness
Anger
Disgust 0.815
Fear 0.754 0.784
Happiness 0.008 0.022 0.066
Reward strategy
(Persuasive-
ness)
0.022 0.041 0.069 0.676
Sadness 0.774 0.801 0.806 0.008 0.032
et al. 2017a; Henseler et al. 2014). In summary, for PLS-SEM model validity and
reliability, the measurement models yielded acceptable values for all indices. Tables 4
and 5show the Cronbach’s alpha, Composite Reliability, AVE, and HTMT values
of the perceived persuasiveness/effectiveness of Reward strategy and the perceived
emotional states.
3.4 Structural model
Our structural models revealed the differences in how people at various stages of
change respond emotionally to each strategy. For each stage of change, we modelled
the relations between each strategy and the five emotional states (see Fig. 4). Next,
we explore for significant structural differences between the models for each stage
of change using the multi-group analysis functionality and procedure in SmartPLS 3
(Matthews 2017; Sarstedt et al. 2011; SmartPLS GmbH, n.d.). This involves calculat-
ing path coefficients (β) which measure the influence of a variable on another, as well
as the significance level (p) of each path coefficient (Hair et al. 2011). The multi-group
analysis establishes significant differences between the models for people at various
stages of change with respect to their βvalues (Henseler et al. 2016; Sarstedt et al.
123
Persuasive strategies and emotional states: towards designing …
2011; SmartPLS GmbH, n.d.). In Table 6, we present the individual path coefficients
(β) and their corresponding significance level (p) obtained from our models for each
stage of change.
4 Results
In this section, we present the results of the relationship between the perceived effec-
tiveness of individual strategies and perceived emotional states for people at different
stages of change. Table 6shows the path coefficients (β) and the associated significance
level (p) obtained from our SmartPLS models.
4.1 Relationship between perceived effectiveness of strategies and perceived
emotional states across different stages of change
The results of our multi-group analysis showed that the perceived effectiveness of
individual strategies is associated with specific perceived emotion(s) at varying degrees
for people at distinct stages of behaviour change, as shown in Table 6.
4.1.1 Self-monitoring strategy
With self-monitoring, individuals are able to assess their progress or performance
towards set goals. In other words, the strategy allows people to track their own
behaviours and observe how far (or close) they are to achieving the desired behaviour.
Based on our results (see Table 6), self-monitoring is significantly associated with
happiness across all the stages of behaviour change, especially Contemplation stage
where the association is strongest (β0.74, p< 0.001). Besides participants at the
Action stage, self-monitoring is significantly associated with at least one negative emo-
tional state for those at other stages of change. Specifically, the strategy is significantly
associated with anger for participants at the Precontemplation stage (β0.33, p<
0.05) and significantly associated with disgust (β0.31, p< 0.05) and sadness (β
0.30, p< 0.05) for those at the Maintenance stage of behaviour change. For those
at the Contemplation, Preparation, and Termination stages of change, self-monitoring
is significantly associated with fear:(β0.32, p< 0.05), (β0.19, p< 0.05), and
(β0.24, p< 0.05), respectively. Below are sample comments from participants to
support our findings:
“Considering the importance of resilience in overcoming life challenges to
achieve my goals, the application will help me focus more on things that matters,
push me towards my goal while I monitor how that battery is increasing daily.
On the interim, I’d also do some other things to be able to see what reduces the
battery and things that increase it. This will help me know things to focus on and
leave behind”. – P5 (Action stage). “I enjoy the idea of tracking moods, mon-
itoring my resilience battery, and breaking my habits up into different sections
123
O. Oyebode et al.
Table 6 Standard path coefficients and significance of the relationships between perceived effectiveness
of individual strategies and perceived emotional states for people at different stages of change. Bolded
coefficients have significance level of p< .001, while unbolded coefficients have significance level of
p< .05. “–” represents non-significant coefficients
Strategy Stage of Change Anger Disgust Fear Happiness Sadness
Self-monitoring Precontemplation .33 – – .47 –
Contemplation – – .32 .74 –
Preparation – – .19 .46 –
Action – – – .43 –
Maintenance – .31 – .48 .30
Termination – – .24 .45 –
Reminder Precontemplation – – – .57 –
Contemplation – – – .71 –
Preparation – – .15 .53 –
Action −.34 −.25 – .61 –
Maintenance – .31 .36 .59 .26
Termination – – .14 .64 –
Expertise Precontemplation – – .36 .62 –
Contemplation – – – .56 –
Preparation – .14 .13 .62 –
Action – – – .55 –
Maintenance – – – .57 –
Termination – – – .67 –
Reduction Precontemplation – – – .53 –
Contemplation – – – .71 –
Preparation – – – .58 –
Action −.35 −.30 −.22 .46 −.29
Maintenance – – – .64 –
Termination – – – .57 –
Rehearsal Precontemplation .29 – – .65 –
Contemplation – – – .80 –
Preparation – – – .62 –
Action −.28 – – .71 –
Maintenance – – .31 .66 –
Termination – – .14 .63 –
Goal setting Precontemplation – – – .52 –
Contemplation – – – .68 –
Preparation −.15 – – .69 –
Action – – – .63 –
Maintenance – – – .66 –
123
Persuasive strategies and emotional states: towards designing …
Table 6 (continued)
Strategy Stage of Change Anger Disgust Fear Happiness Sadness
Termination – – – .65 –
Suggestion Precontemplation – .35 .38 .81 –
Contemplation – – – .78 –
Preparation – – .15 .66 –
Action −.23 – – .75 –
Maintenance – – .29 .69 –
Termination – – .16 .68 –
Reward Precontemplation – – – .78 –
Contemplation – – – .70 –
Preparation – – – .62 –
Action – – – .62 –
Maintenance – – – .72 –
Termination – – – .65 –
Recognition Precontemplation – – – .76 –
Contemplation – −.37 – .76 –
Preparation – – .19 .71 –
Action −.29 ––.80 −.25
Maintenance – – – .63 –
Termination −.23 – – .82 –
Distraction or
Avoidance
Precontemplation – – – .72 –
Contemplation – – – .86 –
Preparation – .13 .18 .60 –
Action – – – .64 –
Maintenance – – – .77 –
Termination – – .17 .74 –
Opportunity to plan for
barriers
Precontemplation – – – .71 –
Contemplation – – – .76 –
Preparation – – – .66 –
Action −.20 – – .63 –
Maintenance – – – .36 –
Termination – – – .61 –
to really examine them”. – P17 (Contemplation stage). “I wouldn’t feel par-
ticularly angry about using the app, neither disgusted...I would be somewhat
afraid…regarding what the results would be…” – P69 (Preparation stage).
123
O. Oyebode et al.
4.1.2 Reminder strategy
Across all the six stages of change, the reminder strategy is significantly and strongly
associated with happiness, based on our findings. This means that participants are
happy with a strategy that reminds them to perform their target behaviour. In addition,
for participants at Precontemplation and Contemplation stages, reminder is signifi-
cantly associated with positive emotion only (i.e. happiness):(β0.57, p< 0.001)
and (β0.71, p< 0.001), respectively, while those in other stages of change expressed
mixed emotions (i.e. both positive and negative emotions) towards the strategy. For
instance, the strategy is significantly associated with both fear and happiness for
participants at the Preparation stage: (β0.15, p< 0.05) and (β0.53, p< 0.001),
respectively, and Termination stage: (β0.14, p< 0.05) and (β0.64, p< 0.001),
respectively. For those at the Action stage, there is a significant association between
reminder strategy and happiness (β0.61, p< 0.001). In addition, the strategy has
significant and negative association with anger (β−0.34, p< 0.001) and disgust
(β−0.25, p< 0.05) which means the intensity of these negative emotional states
(i.e. anger and disgust) decreases instead of increasing. For people at the Maintenance
stage of change, reminder is significantly related to disgust (β0.31, p< 0.05), fear
(β0.36, p< 0.001), and sadness (β0.26, p< 0.05), in addition to happiness (β
0.59, p< 0.001). Sample comments below from participants support our findings:
“Reminders are a good motivation for me because every time I open my phone, I
would see them and I would want to get them done so I can delete the notification
and feel good about it”. – P78 (Precontemplation stage). “At first, I may feel
happy about those reminders and try to do what they say, but couple days later,
if I couldn’t complete some of the reminders, then I may feel afraid to receive
them”. – P99 (Preparation stage).
4.1.3 Expertise strategy
A system that incorporates expertise (i.e. experience, knowledge, and competence)
will have increased persuasive powers (Oinas-Kukkonen and Harjumaa 2009). Partic-
ipants’ emotional response to the expertise strategy tends to reflect this evidence-based
assertion, as the strategy has strong and significant relation with happiness across all
stages of behaviour change (see Table 6). Additionally, there is no significant rela-
tion between expertise and any of the negative emotional states for participants in the
Contemplation, Action, Maintenance, and Termination stages. However, the strategy
is significantly associated with fear for those at the Precontemplation stage (β0.36,
p< 0.05) and with both disgust and fear for participants at the Preparation stage of
change: (β0.14, p< 0.05) and (β0.13, p< 0.05), respectively. Sample comments
below support our findings:
“I love learning new ideas and research, knowing that it is coming from experts
rather than who knows online. I love this as a way to pass on knowledge”. –
P41 (Action stage). “It would be interesting to hear from the experts on how
to improve different areas of my life to become resilient. I believe I would feel
123
Persuasive strategies and emotional states: towards designing …
motivated to include their advice...” – P399 (Contemplation stage). “I feel like
I would be quite bored…It is just other people discussing concepts...” – P33
(Preparation stage).
4.1.4 Reduction strategy
Reducing a complex behaviour (such as resilience building) into simple tasks has
been shown to help users perform the target behaviour (Oinas-Kukkonen and Harjumaa
2009). Our results showed that the reduction strategy has strong and significant relation
with the happiness emotional state across all stages of behaviour change. Interestingly,
happiness is the only emotional state significantly associated with reduction strategy
for participants at Precontemplation (β0.53, p< 0.001), Contemplation (β0.71,
p< 0.001), Preparation (β0.58, p< 0.001), Maintenance (β0.64, p< 0.001),
and Termination (β0.57, p< 0.001) stages of change. For those at the Action stage,
however, the strategy is significantly but negatively associated with anger (β−
0.35, p< 0.001), disgust (β−0.30, p< 0.001), fear (β−0.22, p< 0.05), and
sadness (β−0.29, p< 0.05), hence the strategy reduces these negative emotions
rather than increase them. Below are sample comments from participants to support
our findings:
“I think a simplified menu such as that would make me feel less overwhelmed
and thus more open to exploring the provided techniques and classes and
strategies...” – P54 (Precontemplation stage). “Quick tools are helpful as it
is something that I would use as it is easily accessible to me and therefore I will
complete it more frequently”. – P9 (Action stage).
4.1.5 Rehearsal strategy
The rehearsal strategy is significantly and strongly related to the happiness emotional
state across all the six stages of behaviour change. Our results also showed that there is a
significant association between the strategy and anger for those at Precontemplation
(β0.29, p< 0.05) and Action stages of change, although the relation between
the two constructs is negative but significant for those at the Action stage (β−
0.28, p< 0.05). Moreover, rehearsal strategy is significantly associated with fear for
participants at the Maintenance (β0.31, p< 0.05) and Termination (β0.14, p<
0.05) stages of behaviour change. Sample comments below support our findings:
“I don’t think this would make me feel strongly, but it might be annoying to
practice things before doing them”. – P22 (Precontemplation stage). “I think
actively putting into practice what you have learned really helps to cement
those practices, I like that it guides you through to a certain extent but then
leaves it to you to allow for more freedom”. – P110 (Contemplation stage). “By
providing an area to learn about how to be resilient would make me happy to
know that I am on the right track”. – P165 (Contemplation stage).
123
O. Oyebode et al.
4.1.6 Goal-setting strategy
Goal setting has been shown to increase task motivation and performance, hence effec-
tive for behaviour change (Latham 2020; Lockeand Latham 2002; Strecher et al. 1995).
Our results revealed a strong and significant relation between the goal-setting strat-
egy and happiness across all stages of behaviour change. Also, happiness emerged
as the only emotional state significantly associated with the strategy in all stages of
change, except the Preparation stage in which there exists a significant but negative
association between the strategy and anger (β−0.15, p< 0.05). This negative rela-
tion indicates that the strategy reduces anger emotion instead of increasing it. Hence,
the goal-setting strategy is strongly associated with positive emotion in participants
across all stages of change. Below are sample comments from participants to support
our findings:
“…Personally, I get a lot of pleasure out of setting challenging goals and working
towards them, so I would really enjoy this aspect of the app, and feel like I am
competing with myself in a healthy way”. – P344 (Preparation stage). “I feel
empowered, like this is within my own hands and can be tailored to my unique
circumstances. It would create an immediate urge to dive right in and see what
I need”. – P50 (Contemplation stage).
4.1.7 Suggestion strategy
As regards the suggestion strategy, our results showed that there is a strong and sig-
nificant association between the strategy and the happiness emotional state across
all stages of change. For participants at the Contemplation stage, there is no signif-
icant relationship between the suggestion strategy and the negative emotional states.
However, the strategy is significantly associated with fear for participants at Precon-
templation (β0.38, p< 0.001), Preparation (β0.15, p< 0.05), Maintenance (β
0.29, p< 0.05), and Termination (β0.16, p< 0.05) stages of behaviour change. In
addition, suggestion is significantly associated with disgust (β0.35, p< 0.001) for
participants at the Precontemplation stage, while a significant but negative association
exists between the strategy and anger (β−0.23, p< 0.05) for participants at the
Action stage of change. Below are sample comments from participants to support our
findings:
“I like the idea of getting suggestions based on your current mood. There are
times when I’m feeling down and could use something to make me feel better…”
– P55 (Precontemplation stage). “I feel like I am being monitored and watched.
I would be uneasy about the level of information available to the app and how it
is being used”. – P416 (Precontemplation stage). “This is a great idea! noticing
the users’ mood and suggesting things instead of just bypassing that is perfect.
It makes it feel more personable”. – P331 (Contemplation stage)
123
Persuasive strategies and emotional states: towards designing …
4.1.8 Reward strategy
Our results showed that across all stages of behaviour change, there is a strong and
significant relationship between reward strategy and the happiness emotional state.
In addition, happiness is the only emotional state that is significantly associated with
the strategy across all stages of change (see Table 6). This is unsurprising since reward
is one of the widely used strategies in persuasive systems due to its ability to moti-
vate users by incentivizing them (Orji et al. 2018b). Sample comments revealed why
participants are happy with the reward strategy:
“I’m a very competitive person, and I believe this will be useful for me. I think
this will be motivating for me as I am often driven by reward systems”. – P603
(Precontemplation stage). “I like the idea of being able to earn badges and points
as it creates an incentive for me to go through the same process everyday”. – P322
(Contemplation stage). “Receiving awards for anything I do makes me extremely
happy and keeps me motivated. So the only feeling I would experience from this
is happiness”. – P29 (Preparation stage)
4.1.9 Recognition strategy
Across all stages of behaviour change, recognition strategy has a strong and significant
relationship with happiness emotional state, based on our results. For participants
at the Precontemplation and Maintenance stages, happiness emerged as the only
emotional state that is significantly associated with recognition strategy: (β0.76, p<
0.001) and (β0.63, p< 0.001), respectively. For those at the Action and Termination
stages of change, the strategy is significantly and negatively associated with anger:
(β−0.29, p< 0.001) and (β−0.23, p< 0.05), respectively. Moreover, the
recognition strategy has a significant relationship with fear (β0.19, p< 0.05) for
participants at the Preparation stage of change. For those at the Contemplation stage,
there is a significant but negative association between the strategy and disgust (β
−0.37, p< 0.05), as well as sadness (β−0.25, p< 0.05) for participants at the
Action stage. Below are qualitative comments to justify our findings:
“It feels good to be recognized for hard work so this would motivate me more”. –
P35 (Precontemplation stage). “I also love being noticed and getting attention
for my achievements. This is awesome, it would definitely push me to get my
name up there”. – P98 (Contemplation stage). “I’m sure it feels good to be
publicly recognized. On the other hand, I think I would find it de-motivating if
I were nowhere near the top 10”. – P410 (Preparation stage)
4.1.10 Distraction or avoidance strategy
The distraction or avoidance (DA) strategy provides suggestions and features that
allow users to distract themselves or avoid certain situations while trying to achieve
their target behaviour (McKay et al. 2019). Our results showed that the DA strategy
is strongly and significantly associated with happiness across all the six stages of
123
O. Oyebode et al.
behaviour change. Happiness is also the only emotional state that is significantly
associated with the strategy for participants at Precontemplation (β0.72, p< 0.001),
Contemplation (β0.86, p< 0.001), Action (β0.64, p< 0.001), and Maintenance
(β0.77, p< 0.001) stages of change. However, the DA strategy is significantly
associated with fear (β0.18, p< 0.05) and disgust (β0.13, p< 0.05) for
participants at the Preparation stage of change. The strategy is also significantly related
to fear (β0.17, p< 0.05) for participants at the Termination stage of change. Below
are sample comments to support our findings:
“I like the idea of having things to help distract you or ease your mind. There
are plenty of times where I get upset or in a bad mood, and need something to
keep my mind off of things”. – P520 (Precontemplation stage). “I think I’d be
happy about options to distract away from issues but I worry that it may be too
distracting to the point that you’d rather escape your issue rather than address
it”. – P181 (Preparation stage)
4.1.11 Opportunity to plan for barriers (OPB) strategy
The OPB strategy encourages users to plan and reflect on potential barriers to behaviour
change and identify ways of overcoming them (McKay et al. 2019). Based on our
findings, the strategy is significantly associated with happiness across all stages of
change. In addition, happiness is the only emotional state significantly associated
with the OPB strategy for participants at Precontemplation (β0.71, p< 0.001),
Contemplation (β0.76, p< 0.001), Preparation (β0.66, p< 0.001), Maintenance
(β0.36, p< 0.05), and Termination (β0.61, p< 0.001) stages of change. However,
for participants at the Action stage, the strategy is also significantly but negatively
associated with the anger emotional state (β−0.20, p< 0.05). Below are sample
comments from participants to support our findings:
“I like this strategy alot, because it’s addressing what may be in your way
and giving you help to overcome the challenges”. – P260 (Preparation stage).
“I really like this recognition of the barriers between me and resilience like
achieving a good work-life balance and stressful situations. Very cool”. – P49
(Action stage). “Great idea! I think this would keep the road to success clear in
the user’s mind and understand the speed bumps in their path better by thinking
proactively”. – P391 (Maintenance stage)
4.2 Summary of findings
The results of our analysis showed that each persuasive strategy is significantly asso-
ciated with at least one emotional state across all stages of change (see Table 6). Table
7presents the summary of our results. Specifically, all the eleven strategies are sig-
nificantly associated with happiness across all the six stages of behaviour change;
however, associations with the four negative emotional states (i.e. anger, disgust, fear,
and sadness) differ across the different stages of change.
123
Persuasive strategies and emotional states: towards designing …
Table 7 Mapping persuasive strategies to corresponding emotional states across different stages of change, based on the multi-group analysis
Stage of change Anger Disgust Fear Happiness Sadness
Precontemplation Self-monitoring
Rehearsal
Suggestion Expertise
Suggestion
All strategies –
Contemplation – Recognition Self-monitoring All strategies –
Preparation Goal setting Expertise
Distraction or Avoidance
Self-monitoring
Reminder
Expertise
Suggestion
Recognition
Distraction or Avoidance
All strategies –
Action Reminder
Reduction
Rehearsal
Suggestion
Recognition
Opportunity to plan for barriers
Reminder
Reduction
Reduction All strategies Reduction
Recognition
Maintenance – Self-monitoring
Reminder
Reminder
Rehearsal
Suggestion
All strategies Self-monitoring
Reminder
Termination Recognition – Self-monitoring
Reminder
Rehearsal
Suggestion
Distraction or Avoidance
All strategies –
“–” means that there is no relationship between perceived effectiveness of the strategy and the corresponding perceived emotional state. Strategies highlighted in italics have
negative association with that emotional state
123
O. Oyebode et al.
For instance, for participants at the Precontemplation stage, self-monitoring and
rehearsal strategies are significantly associated with anger, while suggestion strategy
is significantly associated with fear and disgust. In addition, expertise strategy is sig-
nificantly associated with fear. For those at the Contemplation stage, self-monitoring
is significantly associated with fear, while recognition strategy is significantly but
negatively associated with disgust; however, none of the strategies is significantly
associated with anger and sadness. For participants at the Preparation stage, goal-
setting strategy is significantly and negatively associated with anger, while expertise
and distraction or avoidance strategies are significantly associated with disgust,but
none of the strategies is significantly associated with sadness.Also,self-monitoring,
reminder,expertise,suggestion,recognition, and distraction or avoidance strategies
are significantly associated with fear. For participants at the Action stage, reminder,
reduction,rehearsal,suggestion,recognition, and opportunity to plan for barriers
strategies are all significantly and negatively associated with anger, while reminder
and reduction strategies are also significantly and negatively associated with disgust.
Moreover, reduction strategy is significantly and negatively associated with fear and
sadness, while recognition strategy is also significantly and negatively associated
with sadness. For participants at the Maintenance stage, none of the strategies is
significantly associated with anger; however, reminder,rehearsal, and suggestion
strategies are significantly associated with fear.Also,self-monitoring and reminder
strategies are significantly associated with disgust and sadness. Lastly, for partic-
ipants at the Termination stage, recognition strategy is significantly and negatively
associated with anger, while self-monitoring,reminder,rehearsal,suggestion, and
distraction or avoidance strategies are significantly associated with fear; however,
none of the strategies is significantly associated with disgust and sadness.
5 Discussion
Research in the area of personalized and emotion-driven adaptation in persuasive
systems design is emerging but lacks empirical evidence to establish if, how, and
why the perceived effectiveness of persuasive strategies leads to certain emotions in
individuals across different stages of behaviour change. We addressed this gap in our
work, which is an important first step in designing and developing tailored, emotion-
adaptive behaviour change interventions that appeal to a wider audience and tailored to
a specific user group based on their current stage of behaviour change while reinforcing
positive emotion and reducing negative emotions.
5.1 Designing for a wider audience
Our findings showed that all the eleven persuasive strategies are significantly asso-
ciated with the happiness emotional state across all stages of change. Hence, these
strategies are able to promote positive emotion (happiness) as users interact with the
persuasive systems operationalizing them. However, most of these strategies are also
associated with one or more negative emotional states (see Table 6).Basedonour
123
Persuasive strategies and emotional states: towards designing …
findings, out of the eleven strategies used in this empirical study, Reward is the only
strategy that is not associated with any of the negative emotional states (i.e. anger, dis-
gust, fear, and sadness) across all the six stages of change because it induces feelings of
personal accomplishment, provides incentives that increase the urge to achieve more
goals, and offers gamified experience, based on participants’ qualitative comments.
In addition, participants generally liked the idea of being able to unlock additional
contents with the rewards already earned. Moreover, Reduction, Opportunity to plan
for barriers (OPB), and Goal setting strategies are negatively associated with negative
emotional state(s) in only one of the stages of change, while no association exists
with negative emotional states in other stages of change. Specifically, people were
generally happy with the Reduction strategy (including those at the action stage who
experienced reduced negative emotions) because it presents interventions in a less
overwhelming or simple manner. Participants also appreciate gaining quick access
to activities which in turn lessens their cognitive load and improve productivity.The
OPB strategy is associated with happiness for participants (and less anger for those at
the action stage) because it raises awareness of obstacles that could hinder their goal
achievement and by offering concrete steps for overcoming those obstacles serves as a
useful problem-solving approach to challenges in their daily lives. Similarly, the Goal
setting strategy is associated with happiness in individuals across all stages of change
(and less anger for those at the preparation stage) because it tremendously sustains
daily motivation, functioning, and stress coping,aswellasallows people to be in
control of their goals. The strategy also promotes accountability,focus alignment,
and sense of self-mastery, according to participants. Therefore, to appeal to a wider
audience while reducing negative emotions and promoting positive emotion, design-
ers of persuasive systems should employ Reward, Reduction, Opportunity to plan for
barriers, and goal-setting strategies. The Reward strategy could be operationalized or
implemented as points, badges, trophies, ribbons, streak coins, or animated graphics
(e.g. growing tree) that users earn for attempts at completing their tasks. Tangible
rewards (such as cash or gift cards) can also be offered; however, designers should
apply caution to avoid redirecting the actual benefits of adopting a desired behaviour
to the value derived from such rewards (Orji et al. 2014). In addition, designers should
provide a means of exchanging accumulated virtual rewards (such as points) for more
beneficial or exclusive contents that move users closer to their goals. To employ the
Reduction strategy, designers should break a complex behaviour (such as resilience
building) into smaller or simple tasks that are easily achievable, as shown in the fol-
lowing comment: “Having access to simplified content can be very beneficial on busy
days or days that you are not as motivated. Having simpler tasks can encourage users
to maintain their habit of using the app and fostering resiliency without discourag-
ing them”. [P613]. As regards the goal-setting strategy, designers should provide a
means of setting individual goals (which could be daily or weekly) and offer some
level of agency to users in adjusting their goals as necessary. Lastly, to operationalize
the Opportunity to plan for barriers (OPB) strategy, designers could anticipate poten-
tial obstacles that may hinder individual users from completing or achieving their set
goals, and then provide in-app contents, tips, or interventions that help to tackle those
obstacles.
123
O. Oyebode et al.
5.2 Designing for users at different stages of behaviour change
Our results showed that people’s current stage of change can inform effective tailoring
of emotion-adaptive persuasive systems that motivate behaviour change and regulate
users’ emotional states. As shown in Table 6, all the eleven persuasive strategies
are associated with positive emotion (happiness) across all stages of change; however,
strategies associated with the four negative emotional states differ across the six stages
of change. Table 7presents the mapping of persuasive strategies to their corresponding
emotional states across different stages of change. Strategies highlighted in grey colour
may not be avoided by designers since they reduce the negative emotions to which
they are negatively associated.
5.2.1 Precontemplation stage
For people at the Precontemplation stage, persuasive system designers may not employ
Self-monitoring (SM), Rehearsal, Suggestion, and Expertise strategies since they are
associated with negative emotions. However, SM could be important since individuals
at this stage are often unaware that their behaviour is problematic and has negative
consequences (LaMorte 2019; Prochaska et al. 2015); hence, a strategy that raises
awareness about user behaviours will be appropriate. Research shows that SM pro-
motes self-awareness and raises people’s consciousness about their behaviours with
respect to health and wellness (Orji et al. 2018a). Since SM is associated with anger
for individuals at this stage of change, it could benefit from complementary strategies
such as Reminder and Reward which are both associated with happiness for people at
the Precontemplation stage, based on our findings. Combining SM and Reward strate-
gies would not only reinforce positive emotion (happiness) while reducing negative
emotional state (anger), but also enhances the behaviour change potential of SM (Orji
et al. 2018a). In addition, Reminder strategy could motivate users to complete their set
goals, as revealed in the following comment: “I have a hard time sticking to things but
with reminders I may be able to improve my daily resilience…” [P621]. Therefore, we
suggest that persuasive systems designers should consider employing Self-monitoring
in conjunction with Reward and Reminder strategies to motivate (while reinforcing
positive emotion in) individuals at the Precontemplation stage of behaviour change.
5.2.2 Contemplation stage
For people at the contemplation stage, persuasive system designers may not employ
the Self-monitoring strategy due to its association with fear. Recognition strategy,
however, may be employed because of its negative relation with disgust for people at
this stage of change. Yet, users at the contemplation stage are in a state of decisional
balance and should be shifted such that they place more emphasis on the benefits of
changing their unhealthy behaviours. This can be achieved through self-reevaluation
or self-reflection which is a process of change that involves cognitive and affective
reassessment of one’s self-image which, in turn, is capable of moving contemplators to
a stage where they are ready to act (i.e. Preparation stage) (Prochaska et al. 2015). Self-
reflection involves tracking past events or behaviours (e.g. daily journaling or activity
123
Persuasive strategies and emotional states: towards designing …
logging) and aggregated data exploration with visualizations including trend analysis
(Choe et al. 2017) which require the Self-monitoring strategy. Since the strategy leads
to fear in participants due to concern or worry about achieving their goals on a regular
basis, designers should also operationalize Reward and Reminder strategies to enhance
motivation and reinforce positive emotion.
5.2.3 Preparation stage
For people at the preparation stage of behaviour change, designers could employ the
following persuasive strategies: Reduction, Rehearsal, Reward, and Opportunity to
plan for barriers. Goal setting may also be employed since it has negative association
with anger for people at this stage. Furthermore, to move from preparation stage to
the action stage, the “self-liberation” process of change—which refers to commitment
to behaviour change based on the belief that healthy behaviour can be achieved (LaM-
orte 2019)—is recommended (Prochaska et al. 2015). Hence, persuasive strategies
that increase users’ confidence in their ability to achieve target behaviour—such as
Reminder, Self-monitoring, and Suggestion (Oyebode et al. 2021)—are important for
people at this stage of change. However, Reminder is also associated with the fear emo-
tional state because some participants at this stage of change were afraid of receiving
multiple or excessive reminders that could induce anxiety or become overwhelming.
However, a flexible, non-intrusive, and less frequent reminder will allay users’ fears,
as revealed in the following comments: “…Overall, I think reminders are beneficial
but should not be excessive” [P216] and “I am not a person who likes notifications that
interrupt my thoughts…The ability to choose the notifications’ timing may make that
better” [P84]. Self-monitoring and Suggestion strategies are also associated with fear
for participants at the preparation stage of behaviour change, based on our findings.
Some participants expressed concern about completing daily tasks/activities which in
turn could inhibit progress towards their set goals; however, these could be addressed
through regular reminders and reward system to increase attention/focus and to boost
motivation. Also, the fear of being constantly tracked by system and told what to do
rather than being in control of their own situations was expressed by some participants.
They are also afraid that they could be reminded of their undesirable state through the
tips which may further worsen their mood. Therefore, to increase motivation and pro-
mote positive emotion (while reducing negative emotions) in users, persuasive system
designers should also consider utilizing Self-monitoring (complemented with Reward
and Reminder strategies). The reminders should be flexible, less frequent, and non-
intrusive. In addition, designers should offer contextual Suggestions but allow users
to enable/disable those suggestions and should be mindful of the message contents to
avoid emphasizing users’ challenges rather than offering encouragement and support
to overcome the challenges.
5.2.4 Action stage
For people at the action stage of change, persuasive system designers could employ
some or all eleven persuasive strategies including those that have negative relation to
any of the negative emotional states (see Table 7). Nevertheless, according to Luden
123
O. Oyebode et al.
et al., stimulus control,counterconditioning, and helping relationships are needed to
support people at the action stage and transition them to the maintenance stage where
they can sustain their behaviour change (Ludden and Hekkert 2014). “Stimulus con-
trol” leverages reminders and cues to support and encourage the healthy behaviour,
while “counterconditioning” involves learning healthier behaviours that can substi-
tute for the unhealthy behaviour (LaMorte 2019; Prochaska et al. 2015). “Helping
relationships” can be achieved through personalized and helpful tips supportive of
the behaviour change (Ludden and Hekkert 2014). Stimulus control is linked to the
Reminder strategy, while counterconditioning can be achieved using Rehearsal, Sug-
gestion, and Distraction or Avoidance strategies. Helping relationships can also be
achieved through the Suggestion strategy. Based on our findings, Reminder reduces
anger and disgust, while both Rehearsal and Suggestion strategies also reduces anger.
Therefore, we strongly recommend that designers should prioritize Reminder, Sugges-
tion, Rehearsal, and Distraction or Avoidance strategies in persuasivesystems targeting
users at the action stage of behaviour change.
5.2.5 Maintenance and termination stages
For people at the maintenance and termination stages of change, designers of per-
suasive systems may not employ the following strategies: Self-monitoring, Reminder,
Rehearsal, and Suggestion due to their association with negative emotions. In addition,
Distraction or Avoidance (DA) strategy maybe avoided due to its relation with fear for
those at the termination stage of change. Nevertheless, since people at the maintenance
and termination stages have already adopted the target behaviour in their daily lives,
they need to sustain their healthy behaviours through “counterconditioning” (using
Rehearsal, Suggestion, and DA strategies), “stimulus control” (using Reminder strat-
egy), and “helping relationships” (using Suggestion strategy) processes of change
(Connors et al. 2013). According to participants’ qualitative feedback, Rehearsal
strategy provides walkthrough and practical way of learning that ensure they are
accomplishing their goals in the right way; however, it is associated with fear because
sparing extra time for practice is burdensome. Here is one of the participants com-
ment: “Finding the extra time in my day to complete these tasks may bother me, as
it’s one more responsibility to cause stress” [P19]. This issue can be addressed by
allowing users to jump into actual activities with guidance along the way instead of
prior practice sessions, as revealed in the following comment: “I would rather jump
into the activities with guidance as I go rather than ’practice’…” [P177]. Reminder
and Suggestion strategies are associated with negative emotions in individuals due to
similar concerns highlighted in Sect. 5.2.3 which can be addressed via aforementioned
recommendations (see Sect. 5.2.3). Regarding the DA strategy, participants perceived
it as an effective way to cope with daily stressors and achieve their resilience goals
by enabling them to stay positive and on track, keep the mind relaxed, and brighten
up their mood. However, some participants were afraid that the strategy may prevent
them from thinking about the issue and finding solutions. So, designers will have to
support both perspectives by allowing individuals to decide whether or not they want
to engage with DA strategy in an actual application.
123
Persuasive strategies and emotional states: towards designing …
Therefore, designers should consider employing Reminder, Suggestion, Rehearsal,
and DA strategies in persuasive systems targeting users at the maintenance and ter-
mination stages of change to sustain their behaviour change. To mitigate negative
emotions, designers should make practice sessions optional for users and should pro-
vide additional feature that allows them to ask for contextual help as they perform their
target behaviour. Similarly, designers should operationalize the DA strategy such that
users can opt in or out. Finally, designers should adopt the design recommendations
for Reminder and Suggestion strategies highlighted in Sect. 5.2.3 to reinforce positive
emotion.
6 Limitations
First, we used self-reported persuasiveness/effectiveness of strategies implemented
in prototype applications. It is common for HCI researchers (e.g. Jia et al. 2016;
Mulchandani et al. 2022;Orjietal.2018b; Oyebode et al. 2021; Oyebode and Orji
2022; Wais-Zechmann et al. 2018) to assess attitude or perception as a precursor of
actual behaviour, in line with established theories of planned behaviour (Ajzen et al.
1991) and reasoned action (Hale et al. 2002), to develop useful design guidelines.
Although in certain cases, perception may not always reflect actual behaviour; yet, it
is widely acknowledged in the area of persuasive technology (PT) that both explicit
measure (i.e. user perception, self-assessment, or tendency to comply with distinct
persuasive strategies) and implicit measure (actual responses) are effective approaches
to PT design and both have been shown to be effective (Kaptein et al. 2015). According
to Kaptein, “such an explicit approach could be used to tailor persuasive applications
if we have a questionnaire that elicits the tendencies of individual users to comply with
distinct influence principles, we would be able to measure these tendencies apriori,
and adapt the interaction with the user according to the obtained estimates” (Kaptein
et al. 2015). Hence, our findings hold promises for designing PTs to promote actual
behaviour outcome.
Second, emotional responses collected in this study are based on perceived or
imagined reactions to operationalized persuasive strategies. While imagined emo-
tional reactions may not suggest actual experience, research has shown that emotional
imagery is capable of producing feelings that are often congruent with those obtained
in real situations (Cuthbert et al. 1991). In line with this, empirical evidence revealed
that people’s imagined emotional reactions versus actual reactions to stimuli were
nearly identical (Robinson and Clore 2016). Possible explanation for this finding is
that “implicit beliefs about emotions are either representative of, or constitutive of,
online emotional reactions” and “whether implicit beliefs about emotion are derived
from experience or influence experience, or both, the basic point is that there are
theoretical reasons for expecting the type of convergence found here”. (Robinson and
Clore 2016).
Third, we acknowledge that there are many persuasive strategies and emotional
states in the literature; yet, our empirical study established that persuasive strategies
lead to diverse emotions in individuals at distinct stages of behaviour change which
is an important first step towards a broader research in this area. Fourth, although
123
O. Oyebode et al.
we conducted a large-scale study in the context of resilience building, our prototypes
were developed based on well-established and evidence-based principles for designing
persuasive systems across diverse domains; hence, our findings may be applicable to
other health domains. However, further study is needed to establish this.
Finally, our results established relations between perceived effectiveness of per-
suasive strategies and perceived evoked emotions, which may not imply causal
relationships as further study is needed to establish causation while considering other
factors such as demographics and personality traits.
7 Conclusion and future work
In this paper, we explored how to design emotion-adaptive persuasive systems and
effectively tailor them based on users’ current stage of change, having conducted a
large-scale study of 660 participants to investigate if and how individuals including
those at different stages of behaviour change respond emotionally to various persua-
sive strategies and why. Our work contributes to the advancement of knowledge in
the HCI field by underpinning the significance of people’s emotional responses to
persuasive strategies and established that emotional states are important dimensions
for tailoring and selecting appropriate strategies to improve the effectiveness of per-
suasive or behaviour change systems. We also provided empirical insights into the
relation between perceived effectiveness of persuasive strategies and perceived emo-
tional states across different stages of behaviour change using four well-established
theoretical frameworks. To the best of our knowledge, this work is the first to link
emotion theory (from Ekman’s emotion model) with behaviour change theories (from
PSD and ABACUS frameworks) and stages of change theory (from TTM) to find
patterns in people’s emotional responses to persuasive strategies at distinct stages of
change that can inform the choice of strategies to employ in designing and tailoring
emotion-adaptive persuasive systems to motivate behaviour change. As a secondary
objective, we offered qualitative insights to explain why each strategy leads to spe-
cific emotion(s) across different stages of change based on participants’ comments.
Based on our findings, we offered practical guidelines for designing emotion-adaptive
persuasive applications or systems that appeal to a wider audience and tailored to a
specific user group based on their current stage of behaviour change while regulating
users’ emotion (i.e. promoting positive emotion and reducing negative emotions).
As part of future work, we would evaluate the actual effectiveness or persuasiveness
of the strategies and assess users’ actual emotional states via a real emotion-adaptive
persuasive system implemented following recommendations from this work. Specifi-
cally, the system’s effectiveness will be determined through a longitudinal or field study
and subsequent analysis of real-time data including interaction data (which uncover
users’ behavioural, contextual, and emotional patterns), as well as the pre- and post-
study data collected from participants. Also, we would investigate if our findings
targeted at promoting good mental health and well-being (such as resilience building)
could be applied to other behaviours (e.g. physical activity and healthy eating). Finally,
future research could investigate which persuasive strategy leads to which emotional
123
Persuasive strategies and emotional states: towards designing …
response in people belonging to specific demographic or cultural/ethnic groups, as
well as in certain contexts.
Acknowledgements This research was undertaken, in part, thanks to funding from the Canada Research
Chairs Program. We acknowledge the support of the Natural Sciences and Engineering Research Council
of Canada (NSERC) through the Discovery Grant.
Author contributions OO designed and conducted the research study including data collection, data analy-
sis, and manuscript writing. RO also contributedto the study design, data collection, analysis, and manuscript
writing. DS contributed to study design, data collection, and report review. RO supervised the research.
Declarations
Conflict of interest The authors have no conflict of interest to declare that are relevant to the content of this
article.
Appendix
Figures 5,6,7,8,9,10,11,12, and 13 show the prototypes illustrating the self-
monitoring,reminder,goal setting,rehearsal,opportunity to plan for barriers,
Fig. 5 Prototype illustrating the Self-monitoring strategy
123
O. Oyebode et al.
Fig. 6 Prototype illustrating the Reminder strategy
Fig. 7 Prototype illustrating the Goal-setting strategy
123
Persuasive strategies and emotional states: towards designing …
Fig. 8 Prototype illustrating the Rehearsal strategy
expertise,distraction or avoidance,suggestion, and recognition strategies, respec-
tively.
123
O. Oyebode et al.
Fig. 9 Prototype illustrating the
Opportunity to plan for barriers
strategy
123
Persuasive strategies and emotional states: towards designing …
Fig. 10 Prototype illustrating the
Expertise strategy
123
O. Oyebode et al.
Fig. 11 Prototype illustrating the Distraction or Avoidance strategy
123
Persuasive strategies and emotional states: towards designing …
Fig. 12 Prototype illustrating the
Suggestion strategy
123
O. Oyebode et al.
Fig. 13 Prototype illustrating the
Recognition strategy
References
Abdel-Khalek, A.M.: Measuring happiness with a single-item scale. Soc. Behav. Pers. 34(2), 139–150
(2006). https://doi.org/10.2224/SBP.2006.34.2.139
Ajzen, I., et al.: The theory of planned behavior. Organ. Behav. Hum. Decis. Process. 50(2), 179–211 (1991)
Alexander, R., Aragón, O.R., Bookwala, J., Cherbuin, N., Gatt, J.M., Kahrilas, I.J., Kästner, N., Lawrence,
A., Lowe, L., Morrison, R.G., Mueller, S.C., Nusslock, R., Papadelis, C., Polnaszek, K.L., Helene
Richter, S., Silton, R.L., Styliadis, C.: The neuroscience of positive emotions and affect: implications
for cultivating happiness and wellbeing. Neurosci. Biobehav. Rev. 121, 220–249 (2021). https://doi.
org/10.1016/J.NEUBIOREV.2020.12.002
Alslaity, A., Oyebode, O., Fowles, J., Steeves, D., Orji, R., Suruliraj, B., Orji, R., Steeves, D.: Mobile
applications for health and wellness: a systematic review. Proc. ACM Hum. Comput. Interact. 6, 1–29
(2022). https://doi.org/10.1145/3534525
Appel, M., Lugrin, B., Kühle, M., Heindl, C.: The emotional robotic storyteller: on the influence of affect
congruency on narrative transportation, robot perception, and persuasion. Comput. Hum. Behav. 120,
106749 (2021). https://doi.org/10.1016/J.CHB.2021.106749
123
Persuasive strategies and emotional states: towards designing …
Arshad, R., Baig, M.A., Tariq, M., Shahid, S.: Acceptability of persuasive prompts to induce behavioral
change in people suffering from depression. In: IFIP Conference on Human–Computer Interaction
(INTERACT 2019), 11749 LNCS, 120–139. https://doi.org/10.1007/978-3-030-29390-1_7 (2019)
Arteaga, S.M., Kudeki, M., Woodworth, A.: Combating obesity trends in teenagers through persuasive
mobile technology. ACM SIGACCESS Access. Comput. 94(94), 17–25 (2009). https://doi.org/10.
1145/1595061.1595064
Balsamiq. (n.d.). Balsamiq. Rapid, effective and fun wireframing software. Retrieved March 31, 2019, from
https://balsamiq.com/
Bardram, J.E., Frost, M., Szántó, K., Marcu, G.: The MONARCA self-assessment system: a persuasive
personal monitoring system for bipolar patients. In: Proceedings of the 2nd ACM SIGHIT International
Health Informatics Symposium, pp. 21–30 (2012). https://doi.org/10.1145/2110363.2110370
Berking, M., Wupperman, P.: Emotion regulation and mental health: recent findings, current challenges,
and future directions. Curr. Opin. Psychiatry 25(2), 128–134 (2012). https://doi.org/10.1097/YCO.0B
013E3283503669
Boehm, J.K., Chen, Y., Qureshi, F., Soo, J., Umukoro, P., Hernandez, R., Lloyd-Jones, D., Kubzansky, L.D.:
Positive emotions and favorable cardiovascular health: a 20-year longitudinal study. Prev. Med. 136,
106103 (2020). https://doi.org/10.1016/J.YPMED.2020.106103
Braun, M., Schubert, J., Pfleging, B., Alt, F.: Improving driver emotions with affective strategies. Multimodal
Technol. Interact. 3(1), 21 (2019). https://doi.org/10.3390/MTI3010021
Buhrmester, M., Kwang, T., Gosling, S.D.: Amazon’s mechanical Turk: a new source of inexpensive, yet
high-quality data? In: Kazdin, A.E. (ed.) Methodological Issues and Strategies in Clinical Research,
4th edn., pp. 133–139. American Psychological Association, Washington (2015). https://doi.org/10.
1037/14805-009
Canale, N., Vieno, A., Santinello, M., Chieco, F., Andriolo, S.: The efficacy of computerized alcohol
intervention tailored to drinking motives among college students: a quasi-experimental pilot study.
Am. J. Drug Alcohol Abuse 41(2), 183–187 (2015). https://doi.org/10.3109/00952990.2014.991022
Carrera, P., Muñoz, D., Caballero, A.: Mixed emotional appeals in emotional and danger control processes.
Health Commun. 25(8), 726–736 (2010). https://doi.org/10.1080/10410236.2010.521914
Chin, W.W.: The partial least squares approach to structural equation modeling. Mod. Methods Bus. Res.
295(2), 295–336 (1998)
Ching, C.L., Chan, V.L.: Positive emotions, positive feelings and health: a life philosophy. Linguist. Culture
Rev. 4(1), 1–14 (2020). https://doi.org/10.21744/LINGCURE.V4N1.16
Cho, H., Salmon, C.T.: Fear appeals for individuals in different stages of change: intended and unintended
effects and implications on public health campaigns. Health Commun. 20(1), 91–99 (2006). https://
doi.org/10.1207/S15327027HC2001_9
Choe, E.K., Lee, B., Zhu, H., Riche, N.H.: Understanding self-reflection: how people reflect on personal
data through visual data exploration. In: Proceedings of the 11th EAI International Conference on
Pervasive Computing Technologies for Healthcare, pp. 173–182 (2017). https://doi.org/10.1145/3154
862.3154881
Cialdini, R.B.: Harnessing the science of persuasion. Harv. Bus. Rev. 79(9), 72–81 (2001)
Ciman, M., Donini, M., Gaggi, O., Aiolli, F.: Stairstep recognition and counting in a serious Game for
increasing users’ physical activity. Person. Ubiquitous Comput. 20(6), 1015–1033 (2016). https://doi.
org/10.1007/S00779-016-0968-Y
Cohn, M.A., Fredrickson, B.L., Brown, S.L., Mikels, J.A., Conway, A.M.: Happiness unpacked: positive
emotions increase life satisfaction by building resilience. Emotion 9(3), 361–368 (2009). https://doi.
org/10.1037/A0015952
Condori-Fernandez, N., Lopez, F.S.: Using emotions to empower the self-adaptation capability of software
services. In: Proceedings—2017 IEEE/ACM 2nd International Workshop on Emotion Awareness in
Software Engineering, SEmotion 2017, pp. 15–21 (2017). https://doi.org/10.1109/SEMOTION.20
17.8
Connors, G.J., DiClemente, C.C., Velasquez, M.M., Donovan, D.M.: Substance Abuse Treatment and the
Stages of Change: Selecting and Planning Interventions (Second). Guilford Press, New York (2013)
Curtis, K.E., Lahiri, S., Brown, K.E.: Targeting parents for childhood weight management: development of
a theory-driven and user-centered healthy eating app. JMIR Mhealth Uhealth 3(2), e69 (2015). https://
doi.org/10.2196/mhealth.3857
123
O. Oyebode et al.
Cuthbert, B.N., Vrana, S.R., Bradley, M.M.: Imagery: function and physiology. In: Jennings, J.R., Ackles,
P.K., Coles, M.G.H. (eds.) Advances in Psychophysiology: A Research Annual, vol. 4, pp. 1–42.
Jessica Kingsley Publishers, London (1991)
Dillard, J.P., Seo, K.: Affect and persuasion. In: The SAGE Handbook of Persuasion: Developments in
Theory and Practice, pp. 150–166. SAGE Publications Inc (2012). https://doi.org/10.4135/97814522
18410.N10
Ekman, P.: An argument for basic emotions. Cogn. Emot. 6(3–4), 169–200 (1992). https://doi.org/10.1080/
02699939208411068
Ekman, P.: Basic emotions. Handb. Cogn. Emot. 98(45–60), 16 (1999)
Ekman, P., Ekman, E.: Atlas of Emotions (2016). http://atlasofemotions.org/#continents
Elgarf mahaeg, M., Paiva, A., Castellano, G., Peters chpeters, C., Elgarf, M., Calvo-Barajas, N., Peters, C.:
Reward seeking or loss aversion? Impact of regulatory focus theory on emotional induction in children
and their behavior towards a social robot. In: Proceedings of the 2021 CHI Conference on Human
Factors in Computing Systems, pp. 1–11 (2021). https://doi.org/10.1145/3411764
Evers, K.E., Prochaska, J.O., Johnson, J.L., Mauriello, L.M., Padula, J.A., Prochaska, J.M.: A randomized
clinical trial of a population- and transtheoretical model-based stress-management intervention. Health
Psychol. 25(4), 521–529 (2006). https://doi.org/10.1037/0278-6133.25.4.521
Evers, K.E., Paiva, A.L., Johnson, J.L., Cummins, C.O., Prochaska, J.O., Prochaska, J.M., Padula, J., Gök-
bayrak, N.S.: Results of a transtheoretical model-based alcohol, tobacco and other drug intervention
in middle schools. Addict. Behav. 37(9), 1009–1018 (2012). https://doi.org/10.1016/j.addbeh.2012.
04.008
Ferron, M., Massa, P.: Transtheoretical model for designing technologies supporting an active lifestyle.
Proceedings of the Biannual Conference of the Italian Chapter of SIGCHI, pp. 1–8 (2013). https://doi.
org/10.1145/2499149.2499158
Fogg, B.J., Fogg, G.E.: Persuasive Technology: Using Computers to Change What We Think and Do.
Morgan Kaufmann, Cambridge (2003)
Fredrickson, B.L.: What good are positive emotions? Rev. Gen. Psychol. J. Div. Am. Psychol. Assoc. 2(3),
319 (1998). https://doi.org/10.1037/1089-2680.2.3.300
Fredrickson, B.L., Joiner, T.: Positiveemotions trigger upward spirals toward emotional well-being. Psychol.
Sci. 13(2), 172–175 (2002). https://doi.org/10.1111/1467-9280.00431
Frommel, J., Fischbach, F., Rogers, K., Weber, M.: Emotion-based dynamic difficulty adjustment using
parameterized difficulty and self-reports of emotion. In: CHI PLAY 2018—Proceedings of the 2018
Annual Symposium on Computer–Human Interaction in Play, pp. 173–185 (2018a). https://doi.org/
10.1145/3242671.3242682
Frommel, J., Schrader, C., Weber, M.: Towards Emotion-based Adaptive Games: Emotion recognition via
input and performance features. In: CHI PLAY 2018—Proceedings of the 2018 Annual Symposium on
Computer–Human Interaction in Play, 187–199 (2018b). https://doi.org/10.1145/3242671.3242672
Fuller-Tyszkiewicz, M., Richardson, B., Klein, B., Skouteris, H., Christensen, H., Austin, D., Castle, D.,
Mihalopoulos, C., O’Donnell, R., Arulkadacham, L., Shatte, A., Ware, A.: A mobile app-based inter-
vention for depression: end-user and expert usability testing study. JMIR Ment. Health 5(3), e54
(2018). https://doi.org/10.2196/mental.9445
Ganesh, A., Ndulue, C., Orji, R.: PERMARUN: a persuasive game to improve user awareness and self-
efficacy towards secure smartphone behaviour. In: CHI EA ’21: Extended Abstracts of the 2021 CHI
Conference on Human Factors in Computing Systems, pp. 1–7 (2021). https://doi.org/10.1145/3411
763.3451781
Ghandeharioun, A., McDuff, D., Czerwinski, M., Rowan, K.: Towardsunderstanding emotional intelligence
for behavior change chatbots. In: 2019 8th International Conference on Affective Computing and
Intelligent Interaction, ACII 2019, pp. 8–14 (2019). https://doi.org/10.1109/ACII.2019.8925433
Gloria, C.T.,Steinhardt, M.A.: Relationships among positive emotions, coping, resilience and mental health.
Stress. Health 32(2), 145–156 (2016). https://doi.org/10.1002/SMI.2589
Gonzalez-Ramirez, L.P., De La Roca-Chiapas, J.M., Colunga-Rodriguez, C., Preciado-Serrano, M.D.L.,
Daneri-Navarro, A., Pedroza-Cabrera, F.J., Martinez-Arriaga, R.J.: Validation of health behavior and
stages of change questionnaire. Breast Cancer Targets Ther. 9, 199–205 (2017). https://doi.org/10.
2147/BCTT.S129855
Gross, J.J., Levenson, R.W.: Emotional suppression: physiology, self-report, and expressive behavior. J.
Pers. Soc. Psychol. 64(6), 970–986 (1993). https://doi.org/10.1037/0022-3514.64.6.970
123
Persuasive strategies and emotional states: towards designing …
Gross, J.J., Muñoz, R.F.: Emotion regulation and mental health. Clin. Psychol. Sci. Pract. 2(2), 151–164
(1995). https://doi.org/10.1111/J.1468-2850.1995.TB00036.X
Hair, J.F., Ringle, C.M., Sarstedt, M.: PLS-SEM: indeed a silver bullet. J. Mark. Theory Pract. 19(2),
139–152 (2011). https://doi.org/10.2753/MTP1069-6679190202
Hair, J.F., Hult, G.T.M., Ringle, C.M., Sarstedt, M.: A Primer on Partial Least Squares Structural Equation
Modeling (PLS-SEM), 2nd edn. Sage, California (2017a)
Hair, J.F., Jr.,Sarstedt, M., Ringle, C.M., Gudergan, S.P.: AdvancedIssues in Partial Least Squares Structural
Equation Modeling. SAGE Publications, London (2017b)
Hale, J.L., Householder, B.J., Greene, K.L.: The theory of reasoned action. Persuas. Handb. Dev. Theory
Pract. 14, 259–286 (2002)
Hamari, J., Koivisto, J., Pakkanen, T.: Do persuasive technologies persuade? A review of empirical stud-
ies. In: International Conference on Persuasive Technology (PERSUASIVE 2014), 8462 LNCS,
pp. 118–136 (2014). https://doi.org/10.1007/978-3-319-07127-5_11
Hamby, A., Jones, N.: The effect of affect: an appraisal theory perspective on emotional engagement in
narrative persuasion. J. Advert. 51(1), 116–131 (2022). https://doi.org/10.1080/00913367.2021.1981
498/SUPPL_FILE/UJOA_A_1981498_SM2149.DOCX
Haque, M.S., Kangas, M., Jämsä, T.: A persuasive mHealth behavioral change intervention for promoting
physical activity in the workplace: feasibility randomized controlled trial. JMIR Formative Res. 4(5),
e15083 (2020). https://doi.org/10.2196/15083
Hasan, R., Bertenthal, B.I., Hugenberg, K., Kapadia, A.: Your photo is so funny that I don’t mind violating
your privacy by sharing it: effects of individual humor styles on online photo-sharing behaviors. Conf.
Hum. Factors Comput. Syst. Proc. (2021). https://doi.org/10.1145/3411764.3445258
He, H.A., Greenberg, S., Huang, E.M.: One size does not fit all: applying the transtheoretical model to energy
feedback technology design. Proc. SIGCHI Conf. Hum. Fact. Comput. Syst. 2, 927–936 (2010). https://
doi.org/10.1145/1753326.1753464
Henseler, J., Ringle, C.M., Sarstedt, M.: A new criterion for assessing discriminant validity in variance-
based structural equation modeling. J. Acad. Mark. Sci. 43(1), 115–135 (2014). https://doi.org/10.
1007/s11747-014-0403-8
Henseler, J., Ringle, C.M., Sarstedt, M.: Testing measurement invariance of composites using partial least
squares. Int. Mark. Rev. 33(3), 405–431 (2016). https://doi.org/10.1108/IMR-09-2014-0304
Homer, P.M.: When sadness and hope work to motivate charitable giving. J. Bus. Res. 133, 420–431 (2021).
https://doi.org/10.1016/J.JBUSRES.2021.05.018
Horwath, C.C.: Applying the transtheoretical model to eating behaviour change: challenges and opportu-
nities. Nutr. Res. Rev. 12(2), 281–317 (1999). https://doi.org/10.1079/095442299108728965
Hsu, A., Yang, J., Yilmaz, Y., Haque, M.S., Can, C., Blandford, A.: Persuasive technology for overcoming
food cravings and improving snack choices. In: Proceedings of the SIGCHI Conference on Human
Factors in Computing Systems, pp. 3403–3412 (2014). https://doi.org/10.1145/2556288.2557099
Ibanez, L., Roussel, S.: The effects of induced emotions on environmental preferences and behavior: an
experimental study. PLoS ONE 16(9), e0258045 (2021). https://doi.org/10.1371/JOURNAL.PONE.
0258045
Jalowski, M.: Integrating persuasive technology in participatory design workshops: prototypes for partic-
ipant support. In: International Conference on Persuasive Technology (PERSUASIVE 2020), 12064
LNCS, pp. 30–42 (2020). https://doi.org/10.1007/978-3-030-45712-9_3
Jia, Y., Xu, B., Karanam, Y., Voida, S.: Personality-targeted gamification: a survey study on personality
traits and motivational affordances. In: Proceedings of the 2016 CHI Conference on Human Factors
in Computing Systems, pp. 2001–2013 (2016). https://doi.org/10.1145/2858036.2858515
Jicol, C., Wan, C.H., Doling, B., Illingworth, C.H., Yoon, J., Headey, C., Lutteroth, C., Proulx, M.J., Petrini,
K., & Neill, E.O.: Effects of emotion and agency on presence in virtual reality. In: Proceedings of the
2021 CHI Conference on Human Factors in Computing Systems, pp. 1–13 (2021). https://doi.org/10.
1145/3411764.3445588
Johnson, S.S., Paiva, A.L., Cummins, C.O., Johnson, J.L., Dyment, S.J., Wright, J.A., Prochaska, J.O.,
Prochaska, J.M., Sherman, K.: Transtheoretical model-based multiple behavior intervention for weight
management: effectiveness on a population basis. Prev. Med. 46(3), 238–246 (2008). https://doi.org/
10.1016/j.ypmed.2007.09.010
Kaiser, H.F.: A second generation little jiffy. Psychometrika 35(4), 401–415 (1970)
123
O. Oyebode et al.
Kaptein, M., Markopoulos, P., de Ruyter, B., Aarts, E.: Personalizing persuasive technologies: explicit
and implicit personalization using persuasion profiles. Int. J. Hum. Comput. Stud. 77, 38–51 (2015).
https://doi.org/10.1016/j.ijhcs.2015.01.004
Khan, S., Peña, J.: Using persuasive messages to increase engagement with mental health video game apps.
In: Technology and Health, pp. 353–376. Elsevier (2020). https://doi.org/10.1016/b978-0-12-816958-
2.00016-2
Klein, P., Lawn, S., Tsourtos, G., van Agteren, J.: Tailoring of a smartphone smoking cessation app (Kick.it)
for serious mental illness populations: qualitative study. JMIR Hum. Fact. 6(3), e14023 (2019). https://
doi.org/10.2196/14023
Koshy, V., Park, J.S.: We just use what they give us: understanding passenger user perspectives in smart
homes. Conf. Hum. Fact. Comput. Syst. Proc. (2021). https://doi.org/10.1145/3411764.3445598
Krekhov, A., Emmerich, K., Fuchs, J., Krueger, J.H.: Interpolating happiness: understanding the intensity
gradations of face emojis across cultures. In: Conference on Human Factors in Computing System-
s—Proceedings, pp. 1–17 (2022). https://doi.org/10.1145/3491102.3517661
Kupek, E.: Beyond logistic regression: structural equations modelling for binary variablesand its application
to investigating unobserved confounders. BMC Med. Res. Methodol. 6(1), 1–10 (2006). https://doi.
org/10.1186/1471-2288-6-13
LaMorte, W.W.: The transtheoretical model (stages of change). Boston University School of Public Health
(2019). https://sphweb.bumc.bu.edu/otlt/mph-modules/sb/behavioralchangetheories/BehavioralChan
geTheories6.html
Latham, G.: Goal setting: a five-step approach to behavior change. In: Organizational Collaboration, 1st
edn. Routledge, pp. 10–20 (2020). https://doi.org/10.4324/9781315881201-4
Lee, J.E., Lee, D.E., Kim, K., Shim, J.E., Sung, E., Kang, J.H., Hwang, J.Y.: Development of tailored
nutrition information messages based on the transtheoretical model for smartphone application of an
obesity prevention and management program for elementary-school students. Nurs. Res. Pract. 11(3),
247–256 (2017). https://doi.org/10.4162/nrp.2017.11.3.247
Lee, D., Frey, G., Cheng, A., Shih, P.C.: Puzzle walk: a gamified mobile app to increase physical activity
in adults with autism spectrum disorder. In: 2018 10th International Conference on Virtual Worlds
and Games for Serious Applications, VS-Games 2018—Proceedings. https://doi.org/10.1109/VS-GA
MES.2018.8493439 (2018)
Lieffers, J.R.L., Arocha, J.F., Grindrod, K., Hanning, R.M.: Experiences and perceptions of adults accessing
publicly available nutrition behavior-change mobile apps for weight management. J. Acad. Nutr. Diet.
118(2), 229-239.e3 (2018). https://doi.org/10.1016/j.jand.2017.04.015
Lin, H., Lin, Y., Zheng, Y., Liu, Z., Chang, C.: Design, development and randomised controlled trial of a
smartphone application, ‘QinTB’, for smoking cessation in tuberculosis patients: study protocol. BMJ
Open 9(12), e031204 (2019). https://doi.org/10.1136/BMJOPEN-2019-031204
Locke, E.A., Latham, G.P.: Building a practically useful theory of goal setting and task motivation: a 35-year
odyssey. Am. Psychol. 57(9), 705–717 (2002). https://doi.org/10.1037/0003-066X.57.9.705
López Ibáñez, M., Álvarez, N., Peinado, F.: Towards an emotion-driven adaptive system for video
game music. In: International Conference on Advances in Computer Entertainment, 10714 LNCS,
pp. 360–367 (2018). https://doi.org/10.1007/978-3-319-76270-8_25/FIGURES/1
Ludden, G.D.S., Hekkert, P.: Design for healthy behavior: design interventions and stages of
change. In: The Colors of Care: The 9th International Conference on Design & Emotion,
pp. 482–488 (2014). https://research.utwente.nl/en/publications/design-for-healthy-behavior-design-
interventions-and-stages-of-ch
Machuletz, D., Laube, S., Böhme, R.: Webcam covering as planned behavior. In: Proceedings of the 2018
CHI Conference on Human Factors in Computing Systems, pp. 1–13. https://doi.org/10.1145/3173574
(2018)
Mason, W., Suri, S.: Conducting behavioral research on Amazon’s Mechanical Turk. Behav. Res. Methods
44(1), 1–23 (2012). https://doi.org/10.3758/s13428-011-0124-6
Matthews, L.: Applying multigroup analysis in PLS-SEM: a step-by-step process. In: Partial Least Squares
Path Modeling: Basic Concepts, Methodological Issues and Applications, pp. 219–243. Springer
International Publishing (2017). https://doi.org/10.1007/978-3-319-64069-3_10/COVER
McKay, F.H., Slykerman, S., Dunn, M.: The app behavior change scale: creation of a scale to assess the
potential of apps to promote behavior change. JMIR Mhealth Uhealth 7(1), e11130 (2019). https://
doi.org/10.2196/11130
123
Persuasive strategies and emotional states: towards designing …
Mukherjee, A., Dubé, L.: Mixing emotions: the use of humor in fear advertising. J. Consum. Behav. 11(2),
147–161 (2012). https://doi.org/10.1002/CB.389
Mulchandani, D., Alslaity, A., Orji, R.: Exploring the effectiveness of persuasive games for disease pre-
vention and awareness and the impact of tailoring to the stages of change. In: Human–Computer
Interaction, pp. 1–37 (2022). https://doi.org/10.1080/07370024.2022.2057858
Nabi, R.L., Gustafson, A., Jensen, R.: Framing climate change: exploring the role of emotion in generating
advocacy behavior. Sci. Commun. 40(4), 442–468 (2018). https://doi.org/10.1177/1075547018776019
Nandwani, P., Verma, R.: A review on sentiment analysis and emotion detection from text. Soc. Netw. Anal.
Min. 11(1), 1–19 (2021). https://doi.org/10.1007/S13278-021-00776-6/FIGURES/5
Natt och Dag, Y., Mehlig, K., Rosengren, A., Lissner, L., Rosvall, M.: Negative emotional states and negative
life events: consequences for cardiovascular health in a general population. J. Res. 129, 109888 (2020).
https://doi.org/10.1016/J.JPSYCHORES.2019.109888
Ndulue, C., & Orji, R. (2018). STD PONG: changing risky sexual behaviour in Africa through persuasive
games. In: Proceedings of the Second African Conference for Human Computer Interaction: Thriving
Communities, pp. 1–5. https://doi.org/10.1145/3283458.3283463
Noordewier, M.K., Breugelmans, S.M.: On the valence of surprise. Cogn. Emot. 27(7), 1326–1334 (2013).
https://doi.org/10.1080/02699931.2013.777660
ObjectPlanet Inc. (n.d.). Conduct Online Surveys using Opinio. Retrieved January 3, 2020, from http://
www.objectplanet.com/opinio/
Oduor, M., Oinas-Kukkonen, H.: Committing to change: a persuasive systems design analysis of user
commitments for a behaviour change support system. Behav. Inf. Technol. 40(1), 20–38 (2019).
https://doi.org/10.1080/0144929X.2019.1598495
Oinas-Kukkonen, H., Harjumaa, M.: Persuasive systems design: key issues, process model, and system
features. Commun. Assoc. Inf. Syst. 24(1), 96 (2009)
Oppenlaender, J., Lucero, A., Kuosmanen, E., Hosio, S.: Hardhats and bungaloos: comparing crowdsourced
design feedback with peer design feedback in the classroom; hardhats and bungaloos: comparing
crowdsourced design feedback with peer design feedback in the classroom. In: Proceedings of the
2021 CHI Conference on Human Factors in Computing Systems, pp. 1–14 (2021). https://doi.org/10.
1145/3411764
Orji, R., Moffatt, K.: Persuasive technology for health and wellness: state-of-the-art and emerging trends.
Health Inform. J. 24(1), 66–91 (2018). https://doi.org/10.1177/1460458216650979
Orji, R., Mandryk, R.L., Vassileva, J., Gerling, K.M.: Tailoring persuasive health games to gamer type. In:
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 2467–2476
(2013a). https://doi.org/10.1145/2470654.2481341
Orji, R., Vassileva, J., Mandryk, R.L.: LunchTime: a slow-casual game for long-term dietary behavior
change. Pers. Ubiquit. Comput. 17(6), 1211–1221 (2013b). https://doi.org/10.1007/s00779-012-05
90-6
Orji, R., Vassileva, J., Mandryk, R.L.: Modeling the efficacy of persuasive strategies for different gamer
types in serious games for health. User Model. User Adapt. Interact. 24(5), 453–498 (2014). https://
doi.org/10.1007/s11257-014-9149-8
Orji, R., Nacke, L.E., Di Marco, C.: Towards personality-driven persuasive health games and gamified
systems. Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems—CHI
’17, pp. 1015–1027 (2017). https://doi.org/10.1145/3025453.3025577
Orji, R., Lomotey, R., Oyibo, K., Orji, F., Blustein, J., Shahid, S.: Tracking feels oppressive and ‘pun-
ishy’: exploring the costs and benefits of self-monitoring for health and wellness. Digit. Health 4,
205520761879755 (2018a). https://doi.org/10.1177/2055207618797554
Orji, R., Tondello, G.F., Nacke, L.E.: Personalizing persuasive strategies in gameful systems to gamifi-
cation user types. In: Proceedings of the 2018 CHI Conference on Human Factors in Computing
Systems—CHI ’18, pp. 1–14 (2018b). https://doi.org/10.1145/3173574.3174009
Ort, A., Siegenthaler, P., Fahr, A.: How positivelyvalenced health messages can foster information selection:
evidence from two experiments. Front. Commun. 6, 16 (2021). https://doi.org/10.3389/FCOMM.2021.
534496/XML/NLM
Ortony, A., Clore, G.L., Collins, A.: The Cognitive Structure of Emotions. Cambridge University Press,
Cambridge (1990)
Oyebode, O., Orji, R.: Persuasive strategy implementation choices and their effectiveness: towards per-
sonalised persuasive systems. Behav. Inf. Technol. (2022). https://doi.org/10.1080/0144929X.2022.
2112612
123
O. Oyebode et al.
Oyebode, O., Maurya, D., Orji, R.: Nourish your tree! Developing a persuasive exergame for promoting
physical activity among adults. In: 2020 IEEE 8th International Conference on Serious Games and
Applications for Health, SeGAH 2020, pp. 1–7 (2020). https://doi.org/10.1109/SeGAH49190.2020.
9201637
Oyebode, O., Ndulue, C., Mulchandani, D., A. Zamil Adib, A., Alhasani, M., Orji, R.: Tailoring persuasive
and behaviour change systems based on stages of change and motivation. In: Proceedings of the 2021
CHI Conference on Human Factors in Computing Systems, pp. 1–19 (2021). https://doi.org/10.1145/
3411764.3445619
Parrott, W.G.: Emotions in Social Psychology: Essential Readings. Psychology press, London (2001)
Peng, Z., Kim, T., Ma, X.: GremoBot: exploring emotion regulation in group chat. In: Proceedings of the
ACM Conference on Computer Supported Cooperative Work, CSCW, pp. 335–340 (2019). https://
doi.org/10.1145/3311957.3359472
Pfefferbaum, B.J., Reissman, D.B., Pfefferbaum, R.L., Klomp, R.W., Gurwitch, R.H.: Building resilience
to mass trauma events. In: Handbook of Injury and Violence Prevention, pp. 347–358 (2008). https://
doi.org/10.1007/978-0-387-29457-5_19
Philippe, F.L., Lecours, S., Beaulieu-Pelletier, G.: Resilience and positive emotions: examining the role
of emotional memories. J. Pers. 77(1), 139–176 (2009). https://doi.org/10.1111/J.1467-6494.2008.00
541.X
Plutchik, R.:. A general psychoevolutionary theory of emotion. In: Theories of Emotion, pp. 3–33. Academic
Press (1980). https://doi.org/10.1016/B978-0-12-558701-3.50007-7
Plutchik, R.: A psychoevolutionary theory of emotions. Soc. Sci. Inf. 21(4–5), 529–553 (1982). https://doi.
org/10.1177/053901882021004003
Polizzi, C.P., Lynn, S.J.: Regulating emotionality to manage adversity: a systematic review of the relation
between emotion regulation and psychological resilience. Cogn. Ther. Res. 45(4), 577–597 (2021).
https://doi.org/10.1007/S10608-020-10186-1/TABLES/1
Posner, J., Russell, J.A., Peterson, B.S.: The circumplex model of affect: an integrative approach to affective
neuroscience, cognitive development, and psychopathology. Dev. Psychopathol. 17(3), 715 (2005).
https://doi.org/10.1017/S0954579405050340
Prochaska, J.O.: Transtheoretical model of behavior change. Encyclopedia of Behavioral Medicine,
pp. 2266–2270 (2020)
Prochaska, J.O., Velicer, W.F.: The transtheoretical model of health behavior change. Am. J. Health Promot.
12(1), 38–48 (1997). https://doi.org/10.4278/0890-1171-12.1.38
Prochaska, J.O., Diclemente, C.C., Norcross, J.C.: In search of how people change: applications to addictive
behaviors. J. Addict. Nurs. 5(1), 2–16 (1993). https://doi.org/10.3109/10884609309149692
Prochaska, J.O., Redding, C.A., Evers, K.E.: The transtheoretical model and stages of change. In: Health
Behavior: Theory, Research, and Practice, pp. 125–148. Wiley (2015). https://books.google.ca/books?
id=PhUWCgAAQBAJ
Puddephatt, J.-A., Leightley, D., Palmer, L., Jones, N., Mahmoodi, T., Drummond, C., Rona, R.J., Fear,
N.T., Field, M., Goodwin, L.: A qualitative evaluation of the acceptability of a tailored smartphone
alcohol intervention for a military population: information about drinking for ex-serving personnel
(InDEx) app. JMIR Mhealth Uhealth 7(5), e12267 (2019). https://doi.org/10.2196/12267
Purpura, S., Schwanda, V., Williams, K., Stubler, W., Sengers, P.: Fit4Life: the design of a persuasive tech-
nology promoting healthy behavior and ideal weight. In: Conference on Human Factors in Computing
Systems—Proceedings, pp. 423–432 (2011). https://doi.org/10.1145/1978942.1979003
Rezai, L.S., Burns, C.M.: Using cognitive work analysis and a persuasivedesign approach to create effective
blood pressure management systems. Proc. Int. Symp. Hum. Fact. Ergon. Health Care 3(1), 36–43
(2014). https://doi.org/10.1177/2327857914031005
Richardson, G.E., Shadish, W.R., Cook, T.D., Campbell, D.T.: Emotion regulation and resilience: over-
looked connections. Ind. Organ. Psychol. 9(2), 411–415 (2016). https://doi.org/10.1017/IOP.2016.31
Robinson, M.D., Clore, G.L.: Simulation, scenarios, and emotional appraisal: testing the convergence of
real and imagined reactions to emotional stimuli. Pers. Soc. Psychol. Bull. 27(11), 1520–1532 (2016).
https://doi.org/10.1177/01461672012711012
Rodriguez, P., Ortigosa, A., Carro, R.M.: Detecting and making use of emotions to enhance student moti-
vation in e-learning environments. Int. J. Contin. Eng. Educ. Life Long Learn. 24(2), 168–183 (2014).
https://doi.org/10.1504/IJCEELL.2014.060156
Russell, J.A.: A circumplex model of affect. J. Pers. Soc. Psychol. 39(6), 1161–1178 (1980). https://doi.
org/10.1037/H0077714
123
Persuasive strategies and emotional states: towards designing …
Russell, J.A., Mehrabian, A.: Evidence for a three-factor theory of emotions. J. Res. Pers. 11(3), 273–294
(1977). https://doi.org/10.1016/0092-6566(77)90037-X
Saksono, H., Castaneda-Sceppa, C., Hoffman, J., Morris, V., Seif El-Nasr, M., Parker, A.G.: Storywell:
designing for family fitness app motivation by using social rewards and reflection. In: Conference on
Human Factors in Computing Systems—Proceedings, pp. 1–13 (2020). https://doi.org/10.1145/3313
831.3376686
Salovey, P., Rothman, A.J., Detweiler, J.B., Steward, W.T.: Emotional states and physical health. Am.
Psychol. 55(1), 110–121 (2000). https://doi.org/10.1037/0003-066X.55.1.110
Samad, A.M., Mishra, K., Firdaus, M., Ekbal, A.: Empathetic persuasion: reinforcing empathy and persua-
siveness in dialogue systems. In: Findings of the Association for Computational Linguistics: NAACL
2022, pp. 844–856 (2022). https://aclanthology.org/2022.findings-naacl.63/
Samson, A.C., Kreibig, S.D., Soderstrom, B., Wade, A.A., Gross, J.J.: Eliciting positive, negative and mixed
emotional states: a film library for affective scientists. Cogn. Emot. 30(5), 827–856 (2016). https://
doi.org/10.1080/02699931.2015.1031089/SUPPL_FILE/PCEM_A_1031089_SM0547.ZIP
Sarala, S.M., Sharath Yadav, D.H., Ansari, A.: Emotionally adaptive driver voice alert system for advanced
driver assistance system (ADAS) applications. Proc. Int. Conf. Smart Syst. Invent. Technol. ICSSIT
2018, 509–512 (2018). https://doi.org/10.1109/ICSSIT.2018.8748541
Sarstedt, M., Cheah, J.-H.: Partial least squares structural equation modeling using SmartPLS: a software
review. J. Mark. Anal. 7(3), 196–202 (2019). https://doi.org/10.1057/s41270-019-00058-3
Sarstedt, M., Henseler, J., Ringle, C.M.: Multigroup analysis in partial least squares (PLS) path modeling:
alternative methods and empirical results. Adv. Int. Mark. 22, 195–218 (2011). https://doi.org/10.
1108/S1474-7979(2011)0000022012
Sauter, D.: More than happy: the need for disentangling positive emotions. Curr. Dir. Psychol. Sci. 19(1),
36–40 (2010). https://doi.org/10.1177/0963721409359290
Schnall, R., Bakken, S., Rojas, M., Travers, J., Carballo-Dieguez, A.: mHealth technology as a persuasive
tool for treatment, care and management of persons living with HIV. AIDS Behav. 19(2), 81–89 (2015).
https://doi.org/10.1007/s10461-014-0984-8
Schulz, D.N., Candel, M.J., Kremers, S.P.J., Reinwand, D.A., Jander, A., De Vries, H.: Effects of a web-
based tailored intervention to reduce alcohol consumption in adults: randomized controlled trial. J.
Med. Internet Res. 15(9), e206 (2013). https://doi.org/10.2196/jmir.2568
Shank, D.B.: Technology and emotions. In: Handbooks of Sociology and Social Research: Vol. II,
pp. 511–528. Springer Science and Business Media B.V (2014). https://doi.org/10.1007/978-94-017-
9130-4_24/COVER
Shastri, P.C.: Resilience: building immunity in psychiatry. Indian J. Psychiatry 55(3), 224–234 (2013).
https://doi.org/10.4103/0019-5545.117134
Short, C.A., Barnes, S., Carson, J.F., Platt, I.: Happiness as a predictor of resilience in students at a further
education college. J. Furth. High. Educ. 44(2), 170–184 (2018). https://doi.org/10.1080/0309877X.
2018.1527021
SmartPLS GmbH. (n.d.). Product | SmartPLS. Retrieved August 21, 2020, from https://www.smartpls.com/
Soleymani, M., Pantic, M., Pun, T.: Multimodal emotion recognition in response to videos. IEEE Trans.
Affect. Comput. 3(2), 211–223 (2012). https://doi.org/10.1109/T-AFFC.2011.37
Sona Systems Ltd. (n.d.). SONA: cloud-based Subject Pool Software for Universities. Retrieved September
5, 2022, from https://www.sona-systems.com/default.aspx
Spelt, H.A.A., Westerink,J.H.D.M., Frank, L., Ham, J., IJsselsteijn, W.A.: Physiology-based personalization
of persuasive technology: a user modeling perspective. User Model. User Adapt. Interact. 32(1),
133–163 (2022). https://doi.org/10.1007/S11257-021-09313-8
Strecher, V.J., Seijts, G.H., Kok, G.J., Latham, G.P., Glasgow, R., Devellis, B., Meertens, R.M., Bulger,
D.W.: Goal setting as a strategy for health behavior change. Health Educ. Behav. 22(2), 190–200
(1995). https://doi.org/10.1177/109019819502200207
Taurah, S.P., Bhoyedhur, J., Sungkur, R.K.: Emotion-based adaptive learning systems. In: International
Conference on Machine Learning for Networking, 12081 LNCS, pp. 273–286. https://doi.org/10.
1007/978-3-030-45778-5_18/TABLES/5 (2020)
Thomas, R.J., Masthoff, J., Oren, N.: Can I influence You? Development of a scale to measure perceived
persuasiveness and two studies showing the use of the scale. Front. Artif. Intell. 2(24), 1–14 (2019).
https://doi.org/10.3389/frai.2019.00024
123
O. Oyebode et al.
Tiam-Lee, T.J., Sumi, K.: Adaptive feedback based on student emotion in a system for programming practice.
In: International Conference on Intelligent Tutoring Systems, 10858 LNCS, pp. 243–255. https://doi.
org/10.1007/978-3-319-91464-0_24/TABLES/4 (2018)
Tracy, J.L., Randles, D.: Four models of basic emotions: a review of Ekman and Cordaro, Izard, Levenson,
and Panksepp and Watt. Emot. Rev.3(4), 397–405 (2011). https://doi.org/10.1177/1754073911410747
Tugade, M.M., Fredrickson, B.L.: Regulation of positive emotions: emotion regulation strategies that pro-
mote resilience. J. Happiness Stud. 8(3), 311–333 (2007). https://doi.org/10.1007/s10902-006-9015-4
van ’T Riet, J., Ruiter, R.A.C., Werrij, M.Q., Candel, M.J.J.M., de Vries, H.: Distinct pathways to persuasion:
the role of affect in message-framing effects. Eur. J. Soc. Psychol. 40(7), 1261–1276 (2010). https://
doi.org/10.1002/EJSP.722
VanMeter, F., Cicchetti, D.: Resilience. Handb. Clin. Neurol. 173, 67–73 (2020). https://doi.org/10.1016/
B978-0-444-64150-2.00008-3
Wagener, N., Niess, J., Rogers, Y., Schöning, J.: Mood worlds: a virtual environment for autonomous
emotional expression. In: Proceedings of the 2022 CHI Conference on Human Factors in Computing
Systems, pp. 1–16. https://doi.org/10.1145/3491102.3501861 (2022)
Wais-Zechmann, B., Gattol, V., Neureiter, K., Orji, R., Tscheligi, M.: Persuasive technology to support
chronic health conditions: investigating the optimal persuasive strategies for persons with COPD.
International Conference on Persuasive Technology (PERSUASIVE 2018), pp. 255–266. https://doi.
org/10.1007/978-3-319-78978-1_21 (2018)
Wang, G., Suh, A., Kong, H.: Disorder or driver? The effects of Nomophobia on work-related outcomes in
organizations. In: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems,
pp. 1–12. https://doi.org/10.1145/3173574 (2018)
Wilberz, A., Leschtschow, D., Trepkowski, C., Maiero, J., Kruijff, E., Riecke, B.: FaceHaptics: robot arm
based versatile facial haptics for immersive environments. Conf. Hum. Fact. Comput. Syst. Proc. 20,
1–14 (2020). https://doi.org/10.1145/3313831.3376481
Ye, W., Li, Q., Yu, S.: Persuasive effects of message framing and narrative format on promoting COVID-19
vaccination: a study on Chinese college students. Int. J. Environ. Res. Public Health 18(18), 9485
(2021). https://doi.org/10.3390/IJERPH18189485
Yousef, M., Dietrich, T., Torrisi, G.: Positive, negative or both? Assessing emotional appeals effectiveness
in anti-drink driving advertisements. Soc. Mark. q. 27(3), 195–212 (2021). https://doi.org/10.1177/15
245004211025068
Zhang, X., Li, W., Chen, Xu., Lu, S.: MoodExplorer: towards compound emotion detection via smartphone
sensing. Proc. ACM Interact. Mob. Wear. Ubiquit. Technol. 1(4), 1–30 (2018). https://doi.org/10.1145/
3161414
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under
a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted
manuscript version of this article is solely governed by the terms of such publishing agreement and applicable
law.
Oladapo Oyebode is a Ph.D. candidate at Dalhousie University, Canada. His research interests include
Human-Computer Interaction, Persuasive Technology, Adaptive Systems, Affective Computing, Artificial
Intelligence, Digital Health, and Health Informatics. He has published over 40 peer-reviewed research arti-
cles. He is currently designing, developing, and evaluating technologies to tackle health-related (including
mental health) issues.
Darren Steeves is an Adjunct Professor at the School of Health and Human Performance, Dalhousie
University, Canada. His passion is to help people through living his values of collaboration, innovation,
curiosity, and integrity. He has 25-year background in health and performance, consulting with top exec-
utives, Olympic medalists, and world champion athletes.
123
Persuasive strategies and emotional states: towards designing …
Rita Orji is a Canada Research Chair in Persuasive Technology and a Computer Science Professor at
Dalhousie University where she directs Persuasive Computing Lab. Her research is at the intersection of
technology and human behaviour, focusing on investigating user-centered approaches to designing tech-
nologies that improve lives and promote desirable behaviours.
123
A preview of this full-text is provided by Springer Nature.
Content available from User Modeling and User-Adapted Interaction
This content is subject to copyright. Terms and conditions apply.
