Designing for diversity: Dynamic persuasive strategies in mHealth

Designing for diversity: Dynamic persuasive strategies

in mHealth app development

⋆

Aleise H. McGowan

1,*,†

, Scott Sittig

2,∗,†

, Ryan Benton

3,†

, David Bourrie

3,†

, Sriram

Iyengar

4, †

and Aysu Dalogullari

The University of Southern Mississippi, Hattiesburg MS 39429, USA

University of Louisiana at Lafayette, Lafayette LA 70504, USA

University of South Alabama, Mobile AL 36688, USA

The University of Arizona, Tucson, AZ 85721, USA

Abstract

This study examines the impact of persuasive system design (PSD) in mobile health (mHealth) apps,

focusing on how personalized persuasive strategies, based on users’ psychological characteristics can

enhance engagement, behavior change and efficacy. With the ubiquity of mobile devices reshaping

behavior and perspectives, there's a growing need to personalize digital health technologies to

individual users’ characteristics. This approach challenges the conventional 'one size fits all' model,

recognizing the diversity in user needs and motivations. This research employed a multiphase

experimental design, developing and evaluating 25 mHealth app screens using PSD principles. This

involved rigorous prototyping, expert review, and iterative design, ensuring that each screen

effectively incorporated persuasive elements tailored for diverse user groups. The study's findings

highlight the effectiveness of combining primary task support and dialogue support in mHealth

screens to maximize user engagement. Furthermore, the research underscores the importance of

system credibility and social support in persuasive design, although these elements require careful

implementation due to users’ varying perceptions of persuasiveness among users. This work

significantly contributes to the field by providing insights into how digital health technologies can

be optimally designed to cater to the dynamic psychological makeup of users, ultimately enhancing

user engagement with a focus on behavior change.

Keywords

Persuasive Design, User Engagement, User-Centric Design, Digital Health Interventions 1

1. Introduction

The evolution of Behavior Change Systems (BCS) reflects a dynamic journey, from early

behaviorism ideas to the modern combination of psychology, data science, and user-centric

design, revealing a remarkable story of adaptation and creativity. The potential of BCS to

motivate and support individuals is promising; however, there needs to be more explicit

research on behavioral theory and evidence-based solutions [1]. Mobile health (mHealth) apps

have proven highly effective in encouraging positive health-related behavioral changes among

users [2]. Notably, the persuasive system design framework (PSD) plays a crucial role in guiding

BCSS 2024: The 12

International Workshop on Behavior Change Support Systems, April 10, 2024, Wollongong,

Australia.

∗

Corresponding author.

†

These authors contributed equally.

aleise.mcgowan@usm.edu (A. McGowan); scott.sittig@louisiana.edu (S. Sittig)

CEUR

Workshop

Proceedings

ceur-ws.org

ISSN 1613-0073

the use of persuasive technology, especially in mobile health applications [3, 4]. Recently,

mHealth interventions have gained interest as an innovative strategy to combat disease in a

cost-effective manner [5]. Regular evaluation of persuasive features within these mHealth apps

is essential, and the PSD model serves as a valuable tool for accomplishing this task. The benefits

of persuasive technologies in mHealth engagement, capable of educating, convincing, and

encouraging users throughout their behavioral change, particular in health behavior, are

significant and should not be overlooked [2]. The ubiquity of mobile devices has reshaped our

actions and thoughts, making it valuable and relevant for us to learn and understand the

importance of persuasive technology [6]. The word persuasive refers to synonyms like effective,

cueing, convincing, and compelling. At its root, persuasion is simply an attempt to influence

and convince others on various subjects [7].

Given the diverse range of individuals and their unique needs, the "one size fits all" approach

to solutions has been accepted as often counterproductive and outdated. Individuals vary

significantly in their needs, reflecting the diversity and uniqueness of their personal

characteristics and backgrounds [8]. It is important to accommodate cultural differences,

including race, gender, socioeconomic status, age, and sex, instead of applying one-size-fits-all

solutions in an effort to foster behavior change [9]. By embracing diversity and understanding

individual differences, we can achieve better outcomes, increased efficiency in innovation [10],

and better satisfaction [8]. Examining how individual psychological traits like self-efficacy,

health consciousness, and personality types influence user engagement with mHealth screens

highlights the importance of personalized digital health strategies [11].

2. Background

Digital health behavior change interventions have emerged as potent catalysts for positive

behavioral shifts among healthcare professionals, patients, and the public [12]. Mobile health

applications are evolving, with an increasing implementation of persuasive design features

aimed at enhancing behavior change [2]. A wide array of ideas, concepts, and approaches

comprises the behavior change interventions. These include, but are not limited to, the theory

of reasoned action, the theory of planned behavior, the technology of acceptance model, the

self-efficacy theory, the social cognitive theory, the elaboration likelihood model, cognitive

dissonance theory, goal setting theory, and computer self-efficacy [4]. In addition, the majority

of theories on cognitive health outline the possible connections between psychosocial elements

and healthful conduct [13]. Mobile applications target various health behaviors, including

increasing physical activity, smoking cessation, healthy eating, weight loss, and blood pressure

control, among others [14].

A systematic review of controlled trials in mobile health interventions revealed noteworthy

outcomes for therapies delivered via smartphones and tablets [15]. The findings indicate that

intervention participants demonstrated a heightened success rate in changes across a spectrum

of health behaviors and related outcomes. However, the implementation of mobile health

interventions faces significant challenges, such as user engagement, theoretical underpinnings,

pace and efficiency, effectiveness evaluation, regulation, and ethics. To address these problems

and maximize possibilities, a diverse range of academic disciplines, involving behavioral,

computer, and engineering sciences, as well as user-centered design, will need to collaborate

and share their skills [12]. Integrating behavior change interventions into mobile and social

technologies enables real-time, continuous evaluation, offering advantages over outdated, data-

poor, and infrequent methods.

A recent study focusing on computational models of habit formation proposed a method in

health applications for situations where actions are repeated continually; in promoting healthy

lifestyles, one of the obstacles to effective transformation is the process of breaking poor old

habits and developing healthy new ones [16].Previous research and development in persuasive

technology often adapted a one-size-fits-all approach [17]. Considering that individuals have

different motivations and beliefs, there is a need for digital health technology to become more

personalized [10]. In this literature review, we aim to explore the concept of personalization,

understanding diversity across different industries, and the importance of finding solutions that

are tailored to personal needs. One size does not fit all, especially when tailoring Persuasive

Technology (PT) to an individual [9]. Since persuasive systems aim to target behavior and shape

it into a desired behavior [18], understanding individual needs and adapting to various user

characteristics is essential.

Psychologically, consumer engagement improves when a mHealth app aligns with

individual user preferences, sparking interest and fostering sustained interaction and

commitment. This view, as discussed by Tarute et al., emphasizes how the consumers' focus is

drawn to brands or enterprises that resonate with their interests, thereby encouraging

cognitive, emotional, and behavioral engagement [19]. This includes maintaining commitment

and taking corresponding actions. Notably mHealth apps predominantly attract young adults,

with older adults showing less interest. This demographic skew, highlighted by Mustafa et al.

and Askari et al., narrows the consumer age range, since the elderly often show resistance to

adopting health apps [20, 21]. A significant challenge for mHealth apps lies in sustaining or

even initiating consumer interest. Factors contributing to this issue include missing elements

within the apps, lack of enjoyment, confusion regarding usage, and the need to evaluate

multiple apps to find the most effective ones [21].

There exists a paradigm shift in literature where researchers extended the qualifications for

engagement. Researchers began to home in on the principle of user engagement being driven

by the quality of the users/patients experience [22-24]. While others also measured engagement

by the interaction with digital health technologies, often driven by attributes that naturally

evoke interest in the consumer, which was frequently believed to be reflected by behavior

change in the user [25, 26]. Engagement is also seen as a synergized relationship between digital

health technology and the consumer, in which the consumer is fully immersed and aligned with

the activity [27].

User engagement is also driven by user characteristics. For example, emotional (i.e.

motivation) and behavioral (i.e. response to rewards) characteristics are considered driving

factors for the time and energy users are willing to expend [28]. Breaking from previous more

experience-oriented perspectives of engagement, current engagement concepts require the

users to give their undivided attention to the digital health technology [25]. Achieving synergy

between digital health technologies and consumers is often considered the highest form of user

engagement [27]. As smartphones and other conduits for the delivery of digital health

technologies become more ubiquitous, designers are capable of incorporating customization

features to engage users/patients [24].

While the positive influence that persuasion has on changing an individual's attitude and

behavior has been established [29, 30], researchers have contended the need for personalized

systems that address the individual's personality to increase the effectiveness of digital health

technologies [31, 32]. One-size-fits-all digital health technologies that target behavioral change

to improve the user's health often fail because they do not target the psychological traits that

drive an individual’s motivations and behaviors, due in part to the lack of guidance intervention

designers and data scientists with numerous options face [8].

A dynamic personalized approach to the development of persuasive technologies is

imperative as research has shown that strategies that may influence change in an individual

with one type of psychological type may dissuade another individual with a different

psychological type [9]. Our review of the literature revealed a void in the literature with

scientists seeking a more intimate view of the consumer and how they interact with persuasive

principles in order to help guide the design processes. The design process is furthered impaired

by the lack of an understanding of the psychological characteristics of digital health technology

users [33].

The common use of ‘Argumentum Ad Populum’ (“appeal to the majority”) in persuasive

technology becomes less effective over time as it fails to address the dynamic characteristics of

mHealth app users [34, 35]. Persuasive technology features become obsolete over time because

designers do not address the multivariate, dynamic characteristics of mHealth app users [36].

Given the flawed nature of the design process, to leverage the benefits of successfully engaging

the users of digital health technologies, it is desirable that dynamic features driven by user

characteristics are amalgamated into the design process to better serve the context of user

engagement [37, 38]. A methodical approach which intersects dynamic data driven design

facilitated by persuasive technology will allow researchers and designers of persuasive

technologies to predict the persuasive features that will successfully engage users, thus enabling

effective engagement. The following research study aims to address the challenge of designing

persuasive technology-grounded mHealth app screens that reflect the intended persuasive

characteristics of the designer.

With increasing application of user engagement in digital health technologies, the benefits

of enhanced outcomes are increasingly informed by mixed method approaches driven by data

science [39]. The integration of data science and psychological characteristics has led to

significant advances in predicting individual differences and similarities [40]. The use of data

science allows a user’s personality to be leveraged to anticipate his or her potential needs [41].

Contrast mining can identify the significant personality characteristic differences that may lead

to enhanced persuasiveness among groups of users and patients. By using this information,

designers of digital health technologies can establish enhanced guidelines for the

conceptualization of personalized persuasive intervention design for a given group; this, in turn,

would lead to improved engagement of users. The recognition of additional differences will in

turn allow designers of digital health technologies to better engage users and establish

guidelines in each user/patient group which would help in the conceptualization of a

personalized persuasive intervention design.

3. Methodology

3.1. Design process

To identify factors influencing engagement and intent to use a mHealth application, a multiple-

phase experiment was conducted during the Summer 2020. This experiment utilized a survey-

based approach, examining 25 mHealth app screens designed with persuasive principles to

promote physical activity. This study was conducted with the approval of the Institutional

Review Board (IRB) at the University of South Alabama, ensuring adherence to ethical standards

for research involving human participants. This research focuses on the design and validation

of the mHealth app screens. These screens were designed by adhering to the Persuasive System

Design (PSD) categories and principles delineated by Oinas-Kukkonen and Harjumaa [42]. The

screens were all developed with a unifying theme focused on enhancing and promoting physical

exercise. Contrast Mining was used to evaluate the screens, identifying significant differences

in user responses based on the interrelationships of the combinations of persuasive principles

and individual characteristics of the users [43]. Contrast mining is a subarea of data mining that

focuses on finding contrasting patterns that express significant differences in multiple datasets

or classes, often comparing cases with a desired outcome against those with an undesired

outcome [44].

The development process started with creating wireframe prototypes [45] for the mHealth

screens. In this phase, preliminary designs were sketched on paper, with each sheet serving as

a canvas for one mHealth app screen. Each prototype was thoroughly documented.

Documentation included the persuasive system category, the (primary) design principle as from

Oinas-Kukkonen's framework [42], targeted implantation, details about the mockup, and the

mockup name that would be used throughout the study. Table 1 shows an example of the initial

prototype development. Accompanying this documentation was a detailed description of each

screen's features, which was used to guide the assignment of a unique reference name for each

screen. This name was consistently used throughout the phases of questionnaire development

and subsequent analysis. The final step in this phase was the creation of a sketch for each

prototype, ensuring a visual representation of what each mHealth screen would encompass.

In the final development phase, BuildFire [46], a mobile application development tool was

used. Buildfire was used to create digital, high-fidelity prototypes for each mHealth app screen.

These advanced prototypes were instrumental in actualizing the design objectives set forth

during the wireframe stage. After development, still images of the mHealth screens were

captured using an iPhone XS Max. This technique was chosen to ensure that the captured

images would authentically represent the user's experience on a mobile device. The images

were then transferred from the iPhone to a laptop via email, for further analysis.

Table 1

Examples of the Initial Prototype Development Steps

Persuasive

System

Category

Design

Principle

Targeted

Implementation

Mockup Mockup

Name

Primary

Task

Support

Tunneling Guiding people in

a process step-by-

step to meet a

goal

Fitness program with

step by step workout

plan. Once daily/weekly

goals are reached, the

next set of steps are

given.

Burpee-

Squat

System

Credibility

Support

Trustworthiness Application

should appear to

be truthful, fair,

and unbiased

Display information

guaranteeing HIPAA

compliance to reassure

users that information

will not be shared with

3rd party organizations.

HIPAA

3.2. mHealth screen validation

In the final stage of the development process, the mHealth screen prototypes underwent a

rigorous evaluation by two distinguished experts in persuasive technology. This evaluation

used a blind review format for objective and unbiased assessment. The expert panel included

one reviewer with an extensive 12-year background in the field of persuasive technology and

another with 9 years of experience.

Following the expert inspection and blind review, a consultation was held with the expert

review panel where notes and suggestions were reviewed. The review and modification process

continued until the developer and reviewers reached a consensus. The mHealth screens were

iteratively evaluated, modified, and improved following each expert inspection and blind

review.

For the first round, twenty-three mHealth screens were developed: Add, Start, Burpee-Squat,

Increase, Mountain, Target, Trophy, Late, Calories, Dinner Chat, Tracker, About Us, Stories,

Leaderboard, Journal, Partners, Ads, Strategy, CDC, HIPAA, Contact, Before After, and Yoga.

The developer and reviewers identified eleven mHealth screens with conflicting persuasive

technology principles that required modification: Target, Dinner Chat, About Us, Journal,

Partners, Strategy, HIPAA, Contact, Before After, Yoga, and CDC. There were discrepancies in

the intended persuasive technology principle and the identified principle which necessitated

modifications in the mHealth app screen to ensure coherence in the application of persuasive

principles. The revised mHealth screens were resubmitted for review. CDC was dropped during

the first round because the designed persuasive category was not seen by either of the two

reviewers and the category that was identified was seen in another screen.

The Apple mHealth screen was created to replace CDC and submitted with revisions for

round 2. A consensus was reached on the twenty-three mHealth screens during the second

round. Additionally, three paper and high-fidelity prototypes were created for the remaining

persuasive technology principles following the methods stated above. The additional mHealth

screens (SSL, Avatar, and Recreation) were iteratively evaluated, modified, and improved using

expert inspection and blind review methods used during rounds one and two. The iterative

process resulted in twenty-five mHealth screens designed for the questionnaire that were

agreed upon through the blind review process and one mHealth screen prototype being

discarded. The mHealth screen acceptance by round is shown in Table 2. The X indicates which

round the mHealth screen was accepted and N/A indicates that the screen had not been

developed during the particular round.

Table 2

mHealth Screen Acceptance by Round

4. Results

The iterative process yielded twenty-five mHealth screens used in a comprehensive research

questionnaire.

Table 3 presents the final testing iteration, detailing the principles and principle categories

(PT = primary task support, DS = dialogue support, SC = system credibility support and SS =

social support) for each screen. Figure 1 displays two of the final mHealth screens (Start and

Contact) that were developed.

Screen Name Round 1 Round 2 Round 3

Add X

Start X

Burpee-Squat X

Increase X

Mountain X

Target X

Trophy X

Late X

Calories X

Dinner Chat X

Tracker X

About Us X

Stories X

Leaderboard X

Journal X

Partners X

Ads X

Strategy X

Cdc Dropped N/A N/A

Hipaa X

Contact X

Before After X

Yoga X

Apple N/A Replaced Cdc

Ssl N/A N/A X

Avatar N/A N/A X

Recreation N/A N/A X

Table 3

Mobile App Screen Name with Persuasive principles and Categories

Screen Name Principle 1 (Primary) Principle 2 Principle 3

Add (PT) Tailoring (PT) Tunneling

Start (PT) Reduction (PT) Tunneling

Burpee-Squat (PT) Tunneling (PT) Reduction

Increase (DS) Praise

Mountain (PT) Rehearsal (DS) Suggestion

Target (DS) Praise (PT) Personalization

Trophy (DS) Rewards (DS) Praise

Late (DS) Reminders

Calories (DS) Suggestion

Dinner Chat (DS) Social Role (DS) Praise

Tracker (PT) Self-Monitoring

About Us (SC) Expertise (SC)Trustworthiness (SC) Authority

Stories (SS) Recognition (PT) Simulation (DS) Praise

Leaderboard (SS) Competition

Journal (SS) Social Learning (SS) Social Comparison (SC)Social

Facilitation

Partners (SC) Trustworthiness (SC) Expertise (SC) Authority

Ads (SC) Surface Credibility

Strategy (SC) Authority (SC) Expertise

Apple (SC) Verifiability (SC) Expertise (SC) Authority

HIPAA (SC) Trustworthiness (SC) Surface Credibility

Contact (SC) Real-World Feel

Before After (SC) Normative Influence (PT) Simulation

Yoga (SS) Cooperation (DS) Praise (SS)Social

Comparison

SSL (SC)Third-party

Endorsements

(SC) Trustworthiness

Avatar (DS) Similarity (DS) Liking

Figure 1: Sample mHealth screens developed and accepted during review

Among the 25 screens developed, 5 (20%) featured a primary principle from the primary task

support category, 7 (28%) from the dialogue support category, 8 (32%) from the system

credibility support category, and 5 (20%) from the social support category.

Contrast mining, not driven by hypothesis, effectively uncovers strong correlations between

predictors, guiding future research. This method produced a concise set of rules predicting the

persuasiveness of mHealth screens’ primary categories: primary task support, dialogue support,

system credibility support, or social support would be persuasive. The primary task support

scores exceeded the average perceived persuasiveness score, with screens from the system

credibility category closely matching or surpassing the average score.

This study introduces contrast mining as a novel solution to the gap in PSD frameworks’

lack of systematic data driven decisions. Contrast mining offered a multi-layer insight into its

impact on perceived persuasiveness at the screen level. The Primary Task Support category

ranked highest in the weighted perceived persuasiveness bin. This finding contradicts Drozd

[47], who found no significant relationship between primary task support and perceived

persuasiveness. Screens from the System Credibility category achieved mid-level weighted

perceived persuasiveness scores. The findings that primary task support and system credibility

increase perceived persuasiveness are supported by Lehto [48], who found that primary task

support and system credibility both significantly impact perceived persuasiveness directly.

Screens from the Social Support category scored in the lowest perceived persuasiveness bin.

Figure 2 illustrates the bins of perceived persuasiveness rules generated by Contrast Mining.

Figure 2: Contrast mining category results

The findings suggest that practitioners seeking to develop persuasive digital health

technologies should develop screens using techniques in the primary task support or system

credibility categories. Screens that employ techniques from the social support category need to

be “strongly personalized” in order to achieve perceived persuasive-ness as these produced low

perceived persuasiveness scores in our study. The contrast mining findings also suggest

practitioners should use techniques from the dialogue support category when developing digital

health technologies.

Little is known about the impact of psychological characteristics and the combination of

multiple persuasive techniques on perceived persuasiveness. Drozd et al. [47] discovered that

Primary Task Support and Dialogue Support together significantly impacted perceived

persuasiveness. Additional studies that examine the primary and secondary categories are

needed to determine whether or not the combination of additional categories is driving the

perceived persuasiveness.

5. Discussion

Incorporating accurate persuasive design principles into a mHealth design process involves a

detailed and multifaceted approach [42]. This strategy involves seamlessly blending these

principles into the user experience of the digital health technology.

The Primary Task Support category plays a crucial role in simplifying the user's journey by

breaking down complex tasks into simpler, more manageable steps. This was achieved through

the reduction principle (Start), which simplified the steps involved in starting the weight loss

journey, and the tunneling method (Burpee Squat), which guided users step-by-step in the

process of completing an exercise. The design also incorporated tailoring (Add), adapting the

app’s interface and functionality to individual user needs, interests, and personalities, creating

a more customized experience. To enhance engagement and motivation, personalization

(Target) was a key focus, ensuring that users received timely suggestions, praise, and rewards.

Additionally, self-monitoring (Tracker) features were integrated, allowing users to easily track

their progress and performance. The app also included simulation (Before After) and rehearsal

(Mountain), which depicted a video with a coach showing users how to properly perform an

exercise.

Similarly, the Dialogue Support category enriches the interaction between humans and

digital health technologies, rendering the application more captivating and reactive. This

included integrating aspects like praise and rewards (Increase, Trophy), with the app providing

various forms of positive feedback and visual rewards for task completion and reaching health

milestones. In addition, the app included reminders (Late) and tailored suggestions (Calories),

where reminders helped users stay on track with their health goals and personalized

suggestions offered advice based on individual user data. The design also focused on the

principles of similarity (Avatar) and liking (Avatar), ensuring that the app was not only visually

appealing but also relatable to the users.

The System Credibility category aims to build trust and reliability within the system.

Achieving this involved certifying the app’s trustworthiness (Partners), presenting it as honest

and unbiased to the user. To demonstrate expertise (About Us), the app included content from

knowledgeable and reputable health organizations. Surface credibility (Ads) was also a focus,

achieved through a competent and visually credible ad free design that resonated with users.

Additionally, the app provided a real-world feel (Contact) by connecting users with the

organizations and individuals responsible for the content, thereby enhancing the authenticity

of the information provided. Authority (Strategy) was leveraged by incorporating inputs from

recognized experts and authorities in the field such as the CDC and leading weight loss

authorities, thereby bolstering user trust. To further solidify credibility, the app included third-

party endorsements (SSL) and features for verifiability, allowing users to cross-check and

confirm the safety of their health data with external sources.

Finally, the Social Support category leverages social influence to motivate users. This is

accomplished through the social learning (Journal) principle, which allow users to see others

engaging in target behaviors, creating a sense of community and shared goals. Additionally, the

app incorporates social comparison (Journal) principle, enabling users to measure their

performance against that of their peers, which serves as a motivational tool. The use of

normative influence (Before After) is also employed, harnessing peer pressure in a positive way

to encourage desired behaviors, by sharing before and after pictures of users that lost more than

fifty pounds. Social facilitation (Journal) is integrated to give users the feeling of being part of

a collective effort by allowing them to participate alongside others. The mHealth app screens

also tapped into the innate human tendencies towards cooperation (Yoga) and competition

(Leaderboard), encouraging both collaborative and competitive activities which are designed to

increase engagement. Furthermore, the app includes features for recognition (Stories), publicly

acknowledging user achievements, which not only rewards but also motivates users to continue

their health journey.

Each of these categories and principles is interwoven into the design process to create a

persuasive, engaging, and effective digital health technology that resonates with users on

multiple levels, encouraging positive behavior change and sustained engagement.

6. Conclusion

This study aimed to investigate the effectiveness of persuasive design principles in mHealth

applications and how they contribute to user engagement and perceived persuasiveness. This

research advances the field of persuasive technology by focusing on the user-centric design and

validation of mHealth app screens, employing data-driven methods to assess their effectiveness

in engaging users and influencing behavior change. The use of dynamic data-driven capabilities

is important to advancing perceived persuasiveness which has the potential to successfully

engage users of digital health technologies. A significant limitation of this study was the use of

static screens. Developing a fully developed app will allow researchers to evaluate the

engagement of the digital health tool. Running these studies in tandem will allow researchers

to evaluate engagement on both of those to see if higher perceived persuasiveness leads to

higher engagement.

While this study has provided valuable insights into the application of persuasive design

principles in mHealth applications, further clarification is needed regarding the extent of

personalization in these designs. Understanding how various levels of personalization influence

user engagement and the effectiveness of persuasive strategies could significantly enhance the

development of more tailored and impactful digital health interventions. Future research in this

area should explore the nuances of personalization, examining its potential to meet diverse user

needs and preferences more effectively.

References

[1] A. C. King et al., "Effects of Three Motivationally Targeted Mobile Device Applications on

Initial Physical Activity and Sedentary Behavior Change in Midlife and Older Adults: A

Randomized Trial," PLoS One, 11(6): e0156370 (2016). doi: 10.1371/journal.pone.0156370.

[2] J. Matthews, K. T. Win, H. Oinas-Kukkonen, and M. Freeman, "Persuasive Technology in

Mobile Applications Promoting Physical Activity: a Systematic Review," J Med Syst, 40(3):

72 (2016). doi: 10.1007/s10916-015-0425-x.

[3] A. Hussian, A. Mateen, F. Amin, M. A. Abid, and S. Ullah, "Health Monitoring Apps: An

Evaluation of the Persuasive System Design Model for Human Wellbeing," Information,

14(7), (2023). doi: 10.3390/info14070412.

[4] H. Oinas-Kukkonen, "A foundation for the study of behavior change support systems,"

Personal and Ubiquitous Computing, 17(6): 1223-1235, (2012). doi: 10.1007/s00779-012-

0591-5.

[5] S. Sittig, A. McGowan, and S. Iyengar, "Extensive Review of Persuasive System Design

Categories and Principles: Behavioral Obesity Interventions," J Med Syst, 44(7): 128, (2020).

doi: 10.1007/s10916-020-01591-w.

[6] B. J. Fogg, "Creating Persuasive Technologies: An Eight-Step Design Process," presented at

the Proceedings of the 4th International Conference on Persuasive Technology - Persuasive

'09 (2009).

[7] H. A. A. Spelt, J. H. D. M. Westerink, L. Frank, J. Ham, and W. A. Ijsselsteijn, "Physiology-

based personalization of persuasive technology: a user modeling perspective," User

Modeling and User-Adapted Interaction (2022). doi: 10.1007/s11257-021-09313-8.

[8] E. Engl, P. Smittenaar, and S. K. Sgaier, "Identifying population segments for effective

intervention design and targeting using unsupervised machine learning: an end-to-end

guide," Gates Open Res, 3:1503, (2019), doi: 10.12688/gatesopenres.13029.2.

[9] A. M. Abdullahi, K. Oyibo, R. Orji, and A. A. Kawu, "The Influence of Age, Gender, and

Cognitive Ability on the Susceptibility to Persuasive Strategies," Information, 10(11),

(2019). doi: 10.3390/info10110352.

[10] R. Orji, "Exploring the persuasiveness of behavior change support strategies and possible

gender differences," presented at the 2nd International Workshop on Behavior Change

Support Systems, BCSS 2014, Padua; Italy, (2014).

[11] A. McGowan, S. Sittig, D. Bourrie, R. Benton, and S. Iyengar, "The Intersection of

Persuasive System Design and Personalization in Mobile Health: Statistical Evaluation," (in

English), JMIR Mhealth Uhealth, 10(9): e40576, (2022). doi: 10.2196/40576.

[12] S. Michie, L. Yardley, R. West, K. Patrick, and F. Greaves, "Developing and Evaluating

Digital Interventions to Promote Behavior Change in Health and Health Care:

Recommendations Resulting From an International Workshop," J Med Internet Res, 19(6):1-

13, (2017). doi: 10.2196/jmir.7126.

[13] F. Taj, M. C. A. Klein, and A. Van Halteren, "Motivating Machines: The Potential of

Modeling Motivation as MoA for Behavior Change Systems," Information, 13(5), (2022).

doi: 10.3390/info13050258.

[14] C. Ernsting et al., "Using Smartphones and Health Apps to Change and Manage Health

Behaviors: A Population-Based Survey," J Med Internet Res, 19(4):e101, (2017). doi:

10.2196/jmir.6838.

[15] W. T. Riley, D. E. Rivera, A. A. Atienza, W. Nilsen, S. M. Allison, and R. Mermelstein,

"Health behavior models in the age of mobile interventions: are our theories up to the

task?," Transl Behav Med, 1 (1):53-71, (2011). doi: 10.1007/s13142-011-0021-7.

[16] C. Zhang, J. Vanschoren, A. van Wissen, D. Lakens, B. de Ruyter, and I. J. WA, "Theory-

based habit modeling for enhancing behavior prediction in behavior change support

systems," User Model User-adapt Interact, 32(39):389-415, (2022). doi: 10.1007/s11257-022-

09326-x.

[17] A. M. Abdullahi, R. Orji, and A. A. Kawu, "Gender, Age and Subjective Well-Being:

Towards Personalized Persuasive Health Interventions," Information, 10(10), (2019), doi:

10.3390/info10100301.

[18] V. Mylonopoulou, "Design for health behavior change supportive technology," presented

at the Proceedings of the 10th Nordic Conference on Human-Computer Interaction -

NordiCHI '18, 2018.

[19] A. Tarute, S. Nikou, and R. Gatautis, "Mobile application driven consumer engagement,"

Telematics and Informatics, vol. 34, no. 4, pp. 145-156, 2017, doi: 10.1016/j.tele.2017.01.006.

[20] M. Askari, N. S. Klaver, T. J. van Gestel, and J. van de Klundert, "Intention to use Medical

Apps Among Older Adults in the Netherlands: Cross-Sectional Study," J Med Internet Res,

vol. 22, no. 9, p. e18080, Sep 4 2020, doi: 10.2196/18080.

[21] A. S. Mustafa, N. Ali, J. S. Dhillon, G. Alkawsi, and Y. Baashar, "User Engagement and

Abandonment of mHealth: A Cross-Sectional Survey," Healthcare (Basel), vol. 10, no. 2, Jan

24 2022, doi: 10.3390/healthcare10020221.

[22] S. Taki et al., "Assessing User Engagement of an mHealth Intervention: Development and

Implementation of the Growing Healthy App Engagement Index," JMIR Mhealth Uhealth,

vol. 5, no. 6, p. e89, Jun 29 2017, doi: 10.2196/mhealth.7236.

[23] M. Holdener, A. Gut, and A. Angerer, "Applicability of the User Engagement Scale to

Mobile Health: A Survey-Based Quantitative Study," JMIR Mhealth Uhealth, vol. 8, no. 1,

p. e13244, Jan 3 2020, doi: 10.2196/13244.

[24] M. Lalmas, H. O’Brien, and E. Yom-Tov, Measuring User Engagement (Synthesis Lectures

on Information Concepts, Retrieval, and Services). Morgan & Claypool, 2014.

[25] X. Ren, C. Silpasuwanchai, and J. Cahill, "Human-Engaged Computing: the future of

Human–Computer Interaction," CCF Transactions on Pervasive Computing and

Interaction, vol. 1, no. 1, pp. 47-68, 2019, doi: 10.1007/s42486-019-00007-0.

[26] N. Zagalo, "From Experience to Engagement," in Engagement Design: Designing for

Interaction Motivations. Cham: Springer International Publishing, 2020, pp. 11-30.

[27] K. Salehzadeh Niksirat et al., "Approaching Engagement towards Human-Engaged

Computing," presented at the Extended Abstracts of the 2018 CHI Conference on Human

Factors in Computing Systems, 2018.

[28] H. L. O'Brien, J. Arguello, and R. Capra, "An empirical study of interest, task complexity,

and search behaviour on user engagement," Information Processing & Management, vol.

57, no. 3, 2020, doi: 10.1016/j.ipm.2020.102226.

[29] P. MScMed and F. G. BOccTher, "Persuasive technology and behaviour change in parent-

focused eHealth interventions supporting child health: A scoping review protocol," New

Zealand Journal of Physiotherapy, 47(1): 36-48 (2019).

[30] R. Orji and K. Moffatt, "Persuasive technology for health and wellness: State-of-the-art and

emerging trends," Health Informatics J, 24(1):66-91 (2018). doi: 10.1177/1460458216650979.

[31] M. Kaptein, P. Markopoulos, B. de Ruyter, and E. Aarts, "Personalizing persuasive

technologies: Explicit and implicit personalization using persuasion profiles," International

Journal of Human-Computer Studies, vol. 77, pp. 38-51, 2015, doi:

10.1016/j.ijhcs.2015.01.004.

[32] H. J. Wall, C. C. Campbell, L. K. Kaye, A. Levy, and N. Bhullar, "Personality profiles and

persuasion: An exploratory study investigating the role of the Big-5, Type D personality

and the Dark Triad on susceptibility to persuasion," Personality and Individual Differences,

139:69-76, (2019). doi: 10.1016/j.paid.2018.11.003.

[33] M. Tuman and A. Moyer, "Health intentions and behaviors of health app owners: a cross-

sectional study," Psychol Health Med, 24(7):819-826, (2019). doi:

10.1080/13548506.2019.1576911.

[34] E. Anagnostopoulou, E. Bothos, B. Magoutas, J. Schrammel, and G. Mentzas, "Persuasive

technologies for sustainable mobility: State of the art and emerging trends," Sustainability,

10(7):2128, (2018).

[35] R. Orji and R. L. Mandryk, "Developing culturally relevant design guidelines for

encouraging healthy eating behavior," International Journal of Human-Computer Studies,

72(2):207-223, (2014).

[36] H. Oinas-Kukkonen, "Personalization Myopia: A Viewpoint to True Personalization of

Information Systems," presented at the Proceedings of the 22nd International Academic

Mindtrek Conference on - Mindtrek '18, (2018).

[37] J. A. Naslund et al., "Health behavior models for informing digital technology interventions

for individuals with mental illness," Psychiatr Rehabil J, 40(3):325-335, (2017). doi:

10.1037/prj0000246.

[38] D. Spruijt-Metz and W. Nilsen. (2014, July-Sept. 2014) Dynamic Models of Behavior for

Just-in-Time Adaptive Interventions. IEEE Pervasive Computing 13-17.

[39] R. K. Britt, J. Maddox, S. Kanthawala, and J. L. Hayes, "The impact of mHealth

interventions," in Technology and Health, pp. 271-288, (2020).

[40] W. Bleidorn and C. J. Hopwood, "Using Machine Learning to Advance Personality

Assessment and Theory," Pers Soc Psychol Rev, 23(2):190-203, (2019). doi:

10.1177/1088868318772990.

[41] A. Souri, S. Hosseinpour, and A. M. Rahmani, "Personality classification based on profiles

of social networks’ users and the five-factor model of personality," Human-centric

Computing and Information Sciences, 8(1), (2018). doi: 10.1186/s13673-018-0147-4.

[42] H. Oinas-Kukkonen and M. Harjumaa, "Persuasive Systems Design: Key Issues, Process

Model, and System Features," Communications of the Association for Information Systems,

24, (2009). doi: 10.17705/1cais.02428.

[43] Contrast. (2018). GitHub. Accessed: 08/13/2021. [Online]. Available:

https://github.com/parsahosseini/contrast

[44] G. Dong and J. Bailey, Contrast data mining: concepts, algorithms, and applications. CRC

Press, (2012).

[45] E. Rosenzweig, Successful user experience : strategies and roadmaps. Amsterdam: Elsevier,

(2015).

[46] Buildfire Corporation. https://buildfire.com/

[47] F. Drozd, T. Lehto, and H. Oinas-Kukkonen, "Exploring Perceived Persuasiveness of a

Behavior Change Support System: A Structural Model," in Persuasive Technology. Design

for Health and Safety, (Lecture Notes in Computer Science, ch. 14, pp. 157-168, (2012).

[48] T. Lehto, H. Oinas-Kukkonen, and F. Drozd, "Factors affecting perceived persuasiveness of

a behavior change support system," (2012).