Meta-Analysis of Influencing Factors on Quantified Self Using Multi-Model Fusion

Chen Jie
(School of Management, Anhui University, Hefei 230601, China)

Abstract:
[Purpose/Significance] Exploring the influencing factors of quantified self holds important practical significance for improving national health literacy, enhancing individual health awareness, and promoting healthy behaviors. [Method/Process] This study employs a mixed-methods approach combining qualitative and quantitative research. First, it integrates Temporal Self-Regulation Theory (TST), the Fogg Behavior Model (FBM), and the Reflective-Impulsive Model (RIM) to construct the TST-FBM-RIM model. Then, 32 empirical studies related to quantified self were selected from CNKI and Web of Science, identifying 7 dimensions, 13 influencing factors, and 91 independent effect sizes. Finally, using meta-analysis and dividing the process into two stages—motivation formation and motivation translation—the study examines 13 influencing factors of quantified self while analyzing the moderating roles of gender differences and age differences. [Result/Conclusion] Except for social norms, which show no significant relationship, the remaining 12 independent variables all demonstrate positive correlations with quantified self. Gender differences moderate the effects of self-efficacy, perceived value, and social norms on quantified self; age differences moderate the influences of perceived value, perceived ease of use, and perceived usefulness on quantified self.

Keywords: Quantified self; Temporal Self-Regulation Theory (TST); Fogg Behavior Model (FBM); Reflective-Impulsive Model (RIM); Meta-analysis

1.1 Quantified Self

Quantified Self (QS), proposed by Wired magazine editors Wolf and Kevin in 2007, refers to understanding one's body and mind through self-tracking and data recording, also known as self-tracking or life hacking. Wolf emphasized that QS can serve as a mirror to promote self-improvement, self-exploration, self-awareness, and self-cognition. While QS focuses on data processing and has been described across multiple disciplines, it lacks a strict definition. Information science views it as a dynamically growing personal information system that enables individuals to intuitively understand their healthy living states through personal data analysis. Biomedical science categorizes it into macro and micro measurements, identifying abnormalities and treatments through data comparison. Data science treats it as a method for systematically describing individual phenomena using large data corpora to express new levels of health norms in finer granularity. Human sociology considers it a personal social practice encompassing social interaction and institutions, facilitating health data exchange through social communication. Social media research views it as a meaningful personal experience related to healthy living, dependent not only on individual cognition and emotion but also on the existing technological context. Social psychology sees it as a process covering data collection, visualization, and cross-linking to discover behavioral correlations and inspire new health behaviors. Engineering technology treats it as a tool that promotes empowerment by providing new knowledge and greater control, offering individuals unlimited opportunities to enhance their quality of life. Clinical psychology views it as a process of using digital tools and technologies to track and analyze one's body, encouraging individuals to improve self-awareness and optimize personal performance through data.

Based on these perspectives, this study adopts Lupton's definition, conceptualizing "quantified self" as a behavioral pattern shaped by both social interaction and social norms, dedicated to improving individual health literacy through the dynamic process of data collection and reflection.

1.2 Temporal Self-Regulation Theory and Its Application Value in Quantified Self

Temporal Self-Regulation Theory (TST), proposed by Hall and Fong in 2007, is a theoretical framework for explaining individual health behavior patterns [FIGURE:1]. TST integrates research from personality and social psychology, cognitive psychology, behavioral economics, and neuroscience, comprehensively considering temporal dynamics, individual self-regulation, and behavioral tendencies. TST has been widely applied in explaining, predicting, and intervening in health risk behaviors. The theory analyzes both rational and abstract aspects of health intentions while emphasizing the importance of irrational variations in practical applications. For instance, TST reveals that insomnia patients exhibit sleep procrastination due to insufficient self-regulation in resisting the temptation to stay up late, while night-type individuals experience sleep disorders due to circadian misalignment. In predicting health behaviors, TST indicates that heavy episodic drinking largely depends on individual drinking intentions and behavioral tendencies. For intervention, TST emphasizes planning and implementing future behaviors in specific contexts, such as assisting young and middle-aged stroke patients in developing precise rehabilitation exercise plans to enhance execution willingness and cultivate rehabilitation habits.

This study innovatively incorporates TST's temporal perspective to facilitate the development of quantified self time-management strategies. Temporal valence involves subjective assessment of time value in quantified self, while consistency concepts involve maintaining consistent attitudes and behaviors across different time periods. Both factors jointly promote motivation formation—the ability to set goals and expectations in time management. Additionally, ability facilitates behavior formation, and self-regulation promotes reflection and absorption of behavioral feedback to optimize behavior patterns. By comprehensively considering temporal valence, consistency concepts, behavioral ability, and self-regulation, individuals can more flexibly adapt to different contexts to achieve optimal quantified self outcomes.

1.3 Fogg Behavior Model and Its Application Value in Quantified Self

In studying how interactive computing technology influences psychology, Fogg proposed the Fogg Behavior Model (FBM), which states that behavior (B) is determined by three essential elements: motivation (M), ability (A), and prompt (P)—B=MAP. As shown in [FIGURE:2], when all three elements reach certain levels, individuals are more likely to take action. In FBM, motivation and ability do not always develop synchronously. Ability simplifies behavior execution, while motivation manifests in the background. For example, the rapid development of internet and social media provides a suitable environment for misinformation dissemination (ability), while misinformation exploits psychological mechanisms of pleasure or pain (motivation) to promote sharing behavior. Researchers have used FBM to analyze HPV vaccine acceptance among Nigerian caregivers, finding that location and cost (ability) and community misconceptions (motivation) were primary causes of low vaccination rates. The combination of motivation and ability affects contraceptive use, especially after advertisement exposure (prompt). FBM also guides behavior change strategies and persuasive system design, such as enhancing motivation and ability for elderly home monitoring systems to make behavior change signals effective. It informs mental health app design to change behaviors or attitudes, where encouraging messages enhance motivation, intuitive interfaces improve mood tracking ability, and reminders promote sustained use.

This study innovatively integrates FBM's three core dimensions—motivation, ability, and prompt—to guide individuals in clearly understanding their behavior patterns for healthier choices. Motivation refers to the desire to improve health and quality of life through data recording and analysis. Ability encompasses the skills, knowledge, or resources necessary to perform quantified self. Prompts are external or internal signals triggering quantified self behaviors. By incorporating these dimensions, FBM provides a comprehensive analytical perspective for quantified self formation, helping individuals identify obstacles and opportunities to promote sustained practice.

1.4 Reflective-Impulsive Model and Its Application Value in Quantified Self

The Reflective-Impulsive Model (RIM), proposed by Strack and Deutsch in 2004, explores the mechanisms of reflective and impulsive behaviors in social conduct. As shown in [FIGURE:3], individual behavior is regulated by two interacting systems: the reflective system governed by cognition and the impulsive system triggered by episodic experiences. These systems influence each other during decision-making to jointly determine behavioral outcomes. Behavior generation does not always rely on conscious impulse control but can be achieved through strategic cognitive decisions, where episodic experiences directly guide behavior execution while the reflective system provides foundational support for automated control.

According to RIM, health behavior formation is influenced by both conscious control processes and impulsive tendencies. The reflective system generates behavioral decisions and activates corresponding patterns, but behavioral flexibility can lead to impulsive behaviors due to contextual environments, personality traits, and emotional states. For example, alcohol promotes impulsive thinking's influence on eating behavior, and individuals with low self-control act more impulsively. RIM explains clinical communication skill development in medical education and reflective smartphone disengagement (RSD). The impulsive system operates through spreading activation (episodic cues), such as young consumers making impulsive purchasing decisions under peer influence. The impulsive system also correlates with positive emotions, which increase implicit attitude predictive validity, enabling rapid and spontaneous actions that form automated behavior patterns, as seen in problematic social media use where the impulsive system causes continuous attention while the reflective system involves cognitive behavioral control. RIM tendencies also associate with self-regulation, where high self-regulation groups focus more on reflection and low self-regulation groups lean toward impulsivity.

This study innovatively integrates RIM's dual behavior modes to reveal conflicts and contradictions in quantified self practice—the trade-off between immediate gratification and long-term goals that creates behavioral uncertainty. Episodic experience-driven tendencies root in desires for immediate satisfaction and short-term benefits, while reflective influences from cognitive decisions stem from considerations of long-term goals and benefits. This dual mechanism jointly affects decision-making, making quantified self formation more complex. Through RIM's lens, individuals can gain insight into internal conflicts during quantified self pursuit and adjust behavior patterns for better self-management.

1.5 Constructing the TST-FBM-RIM Fusion Model

While TST, FBM, and RIM have been widely introduced into health behavior research, few studies combine these three models comprehensively. Therefore, based on in-depth examination of these models, this study constructs the TST-FBM-RIM fusion model [FIGURE:4] to analyze quantified self influencing factors. By integrating core elements from TST, FBM, and RIM, the model aims to provide more comprehensive theoretical support for health behavior research. The fusion model not only focuses on individuals' self-regulation capabilities within specific timeframes but also considers environmental factors and individuals' reflective and impulsive tendencies. This theoretical framework can more accurately identify health behavior influencing factors and provide scientific evidence for effective health interventions.

1.5.1 Motivation Formation Process Based on Temporal Dimension and Decision-Making

Motivation formation is a complex psychological process encompassing comprehensive evaluation of quantified self, careful decision-making consideration, and complex interplay of multiple psychological factors. Consistency concepts (C), temporal valence (T), and cognitive decision-making (R) constitute the core of the motivation formation stage. As shown in [FIGURE:4], consistency concepts (C) ensure continuity, stability, consistency, and logic in the quantified self process, avoiding self-contradictory and chaotic behavioral decisions. Temporal valence (T) reflects individuals' sensitivity and emphasis on time value, influencing expectations and preferences for behavioral outcomes at different time points. Cognitive decision-making (R), as the rational thinking component, requires individuals to thoroughly collect and analyze relevant information and scientifically evaluate potential behavioral outcomes during motivation formation.

1.5.2 Motivation Translation Process Based on Self-Regulation and Dual-System Support

The translation of motivation into quantified self involves the combined effects of ability (A), prompt (P), episodic experience (I), and self-regulation (S). Ability (A) serves as the foundation for behavioral translation, determining whether individuals possess sufficient knowledge and skills to execute quantified self. When ability is insufficient, strong motivation may fail to translate into action. Prompt (P) triggers behavior by activating motivation and prompting action. Episodic experience (I) regulates and reinforces the process, where positive experiences enhance ability while negative experiences lead to behavior interruption or abandonment. Self-regulation (S) enables individuals to maintain behavioral continuity and stability when facing cognitive development imbalances, ultimately achieving quantified self through self-motivation and adjustment.

2.1 Research Methods and Tools

Meta-analysis was first proposed by Glass in 1976, using statistical methods to integrate multiple studies, improve conclusion accuracy and credibility, support re-analysis of existing research, reduce errors, and enhance stability and generalizability. Comprehensive Meta-Analysis Version (CMA) is a commonly used meta-analysis tool capable of processing over 100 data formats. This study used CMA 3.0 for overall estimation and moderation effect testing.

2.2 Data Sources

Literature data were collected from CNKI and Web of Science databases. In CNKI, the search used TS="quantified self" OR "self-tracking" OR "life hacker" with document type limited to papers,检索时间截至2024年6月1日, yielding 267 Chinese articles. In Web of Science Core Collection, the search used TS="quantified self" OR "self-tracking" OR "self-quantification" OR "life hacker" AND "empirical research" with document type limited to academic journals,检索时间截至2024年6月1日, yielding 562 English articles. After reading titles, abstracts, and full texts, literature was screened according to two criteria: (1) studies must examine quantified self influencing factors with at least one factor; (2) studies must include sample size and correlation coefficients or statistical indicators from which correlations could be calculated. The final sample included 32 articles for meta-analysis (15 Chinese and 17 English) [TABLE:1].

[TABLE:1] shows the literature coding table including authors, publication year, document type, and influencing factors for each study.

3.1 Data Extraction and Coding

This study coded the 32 sample articles, including study population characteristics, sample sizes, and correlation coefficients [TABLE:2]. To reduce manual error, samples were recoded after a 15-day interval. Correlation coefficient r was selected as the effect size; when not provided, β, t, F, or χ² values were converted using the following formulas:
- r = [t²/(t² + df)]^(1/2)
- r = [χ²/(χ² + df)]^(1/2)
- r = 0.98β + 0.05 (β > 0), r = 0.98β - 0.05 (β < 0)

As shown in [TABLE:2], the 32 studies identified 7 dimensions and 13 influencing factors, yielding 91 independent effect sizes with a total sample size of 37,285.

[TABLE:2] presents the data coding table organized by motivation formation stage (consistency concepts, temporal valence, cognitive decision-making) and motivation translation stage (episodic experience, ability, prompt, self-regulation).

3.2 Heterogeneity Test

Heterogeneity testing determined whether random-effects models were appropriate and whether moderation analysis was necessary. As shown in [TABLE:3], except for social norms (I² = 52.375% < 75%), all other 12 factors had I² values exceeding 75%, indicating that random-effects models were appropriate.

3.3 Relationship Strength

Relationship strength refers to the magnitude of correlation coefficients. According to Cohen's criteria, |r| < 0.1 indicates no correlation, 0.1 ≤ |r| < 0.3 weak correlation, 0.3 ≤ |r| < 0.5 moderate correlation, and 0.5 ≤ |r| < 1 strong correlation. Social norms showed no correlation. Moderate correlations were found for self-empowerment, prior norms, perceived value, social support, and privacy concern. Strong correlations were observed for self-efficacy, continuance intention, perceived enjoyment, perceived ease of use, perceived usefulness, social interaction, and subjective norms [TABLE:3].

Sensitivity analysis revealed that removing any single sample produced effect size fluctuations minimal compared to overall effect values, indicating high stability of meta-analysis results [TABLE:3].

3.4.1 Moderating Effect of Gender Distribution

Study samples were categorized by gender distribution into female-dominated and male-dominated groups. Moderation analysis showed that gender distribution positively moderated the relationships between self-efficacy, perceived value, social norms, and quantified self [TABLE:4].

3.4.2 Moderating Effect of Age Distribution

Samples were divided into younger groups (predominantly ≤30 years) and older groups (predominantly >30 years). Age distribution positively moderated the relationships between perceived value, perceived ease of use, perceived usefulness, and quantified self [TABLE:5].

3.5 Publication Bias

Fail-safe Number (Nfs) was used to detect publication bias by comparing Nfs with 5k+10 (where k = number of studies). Except for social norms, all other 12 factors had Nfs values far exceeding five times their original sample sizes, confirming result reliability and excluding significant publication bias [TABLE:6].

4.1.1 Motivation Formation Stage

Self-efficacy, self-empowerment, prior norms, continuance intention, and perceived value all showed significant positive correlations with quantified self. Self-efficacy refers to individuals' confidence and belief in completing tasks or achieving goals, representing capability assessment and success expectations. High self-efficacy individuals tend to set higher goals and demonstrate resilience. Self-empowerment is the ability to autonomously control and manage one's behaviors and decisions when facing challenges; higher self-empowerment leads to more active participation and stronger autonomy in quantified self. Prior norms refer to accumulated behavioral experience and recognition of effort expectations before action; higher prior norms increase quantified self engagement as accumulated experience and knowledge enhance confidence when facing new challenges. Continuance intention reflects persistence in maintaining action despite difficulties; stronger continuance intention enables long-term quantified self practice without giving up after setbacks. Perceived value is the importance individuals assign to activities or goals; higher perceived value increases willingness to invest time and effort in quantified self.

4.1.2 Motivation Translation Stage

Except for the non-significant relationship with social norms, perceived enjoyment, perceived ease of use, perceived usefulness, social support, social interaction, privacy concern, and subjective norms all showed significant positive correlations with quantified self [FIGURE:5].

Social norms emphasize perceived social status or widely accepted standards. In quantified self, personal decisions primarily depend on intrinsic motivation and cognitive evaluation rather than merely meeting social expectations. Perceived enjoyment refers to the pleasure and fun experienced during quantified self; higher enjoyment leads to more active participation. Perceived ease of use emphasizes perceived technology difficulty, while perceived usefulness emphasizes perceived improvement in personal capabilities. When individuals believe quantified self tools are easy to master and can enhance abilities (e.g., improving habits or efficiency), they engage more actively. Social support, representing respect and social identity in social networks, significantly influences user behavior and creates positive cycles where individual success inspires broader community participation. Social interaction refers to expanding information channels and maintaining social connections through quantified self; encouragement and feedback from social networks reinforce quantified self value and commitment. Privacy concern reflects worries about personal information collection, use, and sharing; high privacy concerns may hinder motivation translation into action. Subjective norm represents the degree to which significant others believe one should perform the behavior; positive social influence strengthens determination and facilitates motivation translation.

4.2.1 Moderating Effect of Gender

Gender differences influence self-efficacy, perceived value, and compliance with social norms in quantified self. In gender research, the relationship between self-efficacy and gender is significant. According to gender theory, women face more gender role constraints and stereotypes that affect their self-efficacy. During quantified self, women may doubt their abilities due to traditional gender roles, particularly in data and technology fields, while men show greater confidence. Perceived value is also gendered: women unfamiliar with technology may perceive lower value in quantified self, while men's easier technology adaptation leads to higher perceived value. Social norms reinforce gender role stereotypes; conservative views toward women make it harder for them to overcome low self-efficacy, hindering active participation. Research on gender differences in academic output shows female faculty face disadvantages in human capital and institutional resources while bearing greater family responsibilities, limiting time and energy for research.

4.2.2 Moderating Effect of Age

Age differences moderate the effects of perceived value, perceived ease of use, and perceived usefulness on quantified self. Age significantly shapes perceived value of quantified self outcomes. Younger groups demonstrate high technology acceptance, quickly adapting to tools and appreciating their convenience and utility. Older groups face challenges in technology adoption and proficiency, reducing positive perceptions of ease of use and usefulness. Research shows Chinese highly-cited chemists are younger than their American counterparts, indicating younger scientists' higher activity and innovation. Age also positively correlates with privacy disclosure behavior: younger users share personal data more readily for personalized services, while older users are more cautious due to different privacy risk perceptions and technology trust levels. Additionally, age affects data interpretation and application—younger users leverage data for self-reflection and adjustment, while older users prefer intuitive data presentation.

5.1 Research Value

First, this study innovatively developed the TST-FBM-RIM model, revealing the mechanisms of motivation formation and translation in quantified self. This provides a new theoretical perspective for understanding how individuals improve health through quantified means, enriching quantified self research and offering solid theoretical support for future empirical studies. Second, the study identified 7 dimensions with 13 influencing factors, using meta-analysis to verify their association strengths with quantified self, providing clearer research pathways and theoretical references for scholars.

5.2 Research Prospects

First, create highly consistent time value to facilitate quantified self motivation and enhance health awareness across social groups. Through education and publicity, improve public understanding of health technology. Use case studies and story sharing to demonstrate how quantified self enables concrete life changes. Establish online communities and forums for experience sharing and mutual support. Encourage users to share achievements via social media to obtain support from friends and family. Second, create quality behavioral environments to achieve quantified self and improve population health. Focus on enhancing data security and privacy protection for health quantified self tools while simplifying user interface design. Provide clear guidance to help users quickly master functions for easy health data tracking and management. Offer data visualization and reporting to help users understand health status and trends. Encourage sharing health data with family and friends to strengthen social support and interaction.

5.3 Research Limitations

First, the empirical data scope is relatively limited; despite meta-analysis integration, result generalizability requires further validation with more data. Second, when examining moderating variables, cultural background and education level impacts were not fully considered; future research should explore these potential moderators to expand perspectives. Finally, understanding of interactions and dynamic relationships among influencing factors remains insufficient. Future studies could employ more complex statistical methods like time series analysis or structural equation modeling to explore these complex relationships and provide deeper insights.

References
[1] Chinese Government Network. Notice on Launching the Three-Year Action for National Health Literacy Improvement (2024-2027) [EB/OL]. [2024-06-09]. https://www.gov.cn/zhengce/zhengceku/202406/content_6955867.html.
[2] Wolf G I. Quantified self [EB/OL]. [2024-01-04]. https://www.webcitation.org/66TEHdz4d?url=http://aether.com/quantifiedself.
[3] Kevin K. What is the quantified self? [EB/OL]. [2024-01-04]. https://www.webcitation.org/66TEY49wv?url=http://quantifiedself.com/2007/page/3//.
[4] Wolf G I. The quantified self [EB/OL]. [2023-12-15]. https://www.ted.com/talk/gary_wolf_the_quantified_self.
[5] Ruckenstein M, Schüll N D. The datafication of health [J]. Annual Review of Anthropology, 2017, 46(1): 261-278.
[6] Li I, Dey A, Forlizzi J. A stage-based model of personal informatics systems [C]//CHI'10: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. 2010: 557-566.
[7] Smarr L. Quantifying your body: a how-to guide from a systems biology perspective [J]. Biotechnology Journal, 2012, 7(8): 980-991.
[8] Swan M. The quantified self: fundamental disruption in big data science and biological discovery [J]. Big Data, 2013, 1(2): 85-99.
[9] Lupton D. Self-tracking cultures: towards a sociology of personal informatics [C]//OzCHI'14: Proceedings of the 26th Australian Computer-Human Interaction Conference on Designing Futures: The Future of Design. 2014: 77-86.
[10] Lomborg S, Franden K. Self-tracking as communication [J]. Information, Communication & Society, 2016, 19(7): 1015-1027.
[11] Maltseva K, Lutz C. A quantum of self: a study of self-quantification and self-disclosure [J]. Computers in Human Behavior, 2018, 81: 102-114.
[12] De Moya J F, Pallud J. From panopticon to heautopticon: a new form of surveillance introduced by quantified-self practices [J]. Information Systems Journal, 2020, 30(6): 940-976.
[13] Findeis C, Salfeld B, Voigt S, et al. Quantifying self-quantification: a statistical study on individual characteristics and motivations for digital self-tracking in young-and middle-aged adults in Germany [J]. New Media & Society, 2023, 25(9): 2300-2320.
[14] Hall P A, Fong G T. Temporal self-regulation theory: a model for individual health behavior [J]. Health Psychology Review, 2007, 1(1): 6-52.
[15] Hall P A, Fong G T. Temporal self-regulation theory: looking forward [J]. Health Psychology Review, 2010, 4(2): 83-92.
[16] Chen C, Zhai Q H, Wang L M, et al. Application progress of temporal self-regulation theory in health behavior [J]. Chinese Journal of Health Education, 2023, 39(12): 1131-1135.
[17] Wang W Q, Wu M J, Hua Y, et al. Application status and prospects of temporal self-regulation theory in health behavior [J]. Journal of Nursing, 2023, 30(4): 43-48.
[18] Fu Y M, Ma X H, Mu L. Analysis of psycho-physiological mechanisms behind sleep procrastination behavior [J]. Psychological Science, 2020, 43(5): 1234-1240.
[19] Black N, Mullan B, Sharpe L. Predicting heavy episodic drinking using an extended temporal self-regulation theory [J]. Addictive Behaviors, 2017, 73: 111-118.
[20] Zhang R, Chen Y P, Huai P P, et al. Effect of implementation intention strategies based on temporal self-regulation theory on rehabilitation exercise intention in young and middle-aged stroke patients [J]. Chinese Nursing Research, 2023, 37(18): 3368-3373.
[21] Fogg B J. Fogg behavior model [EB/OL]. [2024-01-04]. https://behaviormodel.org.
[22] Guimaraes M, Emmendorfer L, Adamatti D. Persuasive agent based simulation for evaluation of the dynamic threshold line and trigger classification from the Fogg Behavior Model [J]. Simulation Modelling Practice and Theory, 2018, 83: 18-35.
[23] Muriel-Torrado E, Pereira D B. Correlations between the concepts of disinformation and Fogg's Behavior Model [J/OL]. Transinformação, 2020, 32(2): e200026. [2025-01-04]. https://doi.org/10.1590/2318-0889202032e200026.
[24] Agha S, Bernard D, Francis S, et al. Determinants of human papillomavirus vaccine acceptance among caregivers in Nigeria: a Fogg Behavior Model-Based approach [J]. Vaccines, 2024, 12(1): 84.
[25] Agha S, Tollefson D, Paul S, et al. Use of the Fogg Behavior Model to assess the impact of a social marketing campaign on condom use in Pakistan [J]. Journal of Health Communication, 2019, 24(3): 284-292.
[26] Chatterjee S, Byun J, Dutta K, et al. Designing an Internet-of-Things(IoT) and sensor-based in-home monitoring system for assisting diabetes patients: iterative learning from two case studies [J]. European Journal of Information Systems, 2018, 27(6): 670-685.
[27] Azar K M J, Lesser L I, Laing B Y, et al. Mobile applications for weight management: theory-based content analysis [J]. American Journal of Preventive Medicine, 2013, 45(5): 583-589.
[28] Alqahtani F, Al Khalifah G, Oyebode O, et al. Apps for mental health: an evaluation of behavior change strategies and recommendations for future development [J/OL]. Frontiers in Artificial Intelligence, 2019, 2. https://doi.org/10.3389/frai.2019.00030.
[29] Strack F, Deutsch R. Reflective and impulsive determinants of social behavior [J]. Personality and Social Psychology Review, 2004, 8(3): 220-247.
[30] Vohs K D. Self-regulatory resources power the reflective system: evidence from five domains [J]. Journal of Consumer Psychology, 2006, 16(3): 217-223.
[31] Eder A B. Control of impulsive emotional behaviour through implementation intentions [J]. Cognition and Emotion, 2011, 25(3): 478-489.
[32] Hofmann W, Friese M, Wiers R W. Impulsive versus reflective influences on health behavior: a theoretical framework and empirical review [J]. Health Psychology Review, 2008, 2(2): 111-137.
[33] Wouda J C, van de Wiel H B M. Education in patient-physician communication: how to improve effectiveness? [J]. Patient Education and Counseling, 2013, 90(1): 46-53.
[34] Matthes J, Karsay K, Hirsch M, et al. Reflective smartphone disengagement: conceptualization, measurement, and validation [J/OL]. Computers in Human Behavior, 2022, 128: 107078. [2025-01-04]. https://doi.org/10.1016/j.chb.2021.107078.
[35] Thürmer Lukas J, Bieleke M, Wieber F, et al. If-then plans help regulate automatic peer influence on impulse buying [J]. European Journal of Marketing, 2020, 54(9): 2079-2105.
[36] Prestwich A, Hurling R, Baker S. Implicit shopping: attitudinal determinants of the purchasing of healthy and unhealthy foods [J]. Psychology & Health, 2011, 26(7): 875-885.
[37] Cheval B, Sarrazin P, Pelletier L. Impulsive approach tendencies towards physical activity and sedentary behaviors, but not reflective intentions, prospectively predict non-exercise activity thermogenesis [J/OL]. PLoS One, 2014, 9(12): e115238. [2025-01-04]. https://doi.org/10.1371/journal.pone.0115238.
[38] Turel O, Qahri-Saremi H. Problematic use of social networking sites: antecedents and consequence from a dual-system theory perspective [J]. Journal of Management Information Systems, 2016, 33(4): 1087-1116.
[39] Wien A H. Self-presentation via electronic word of mouth-a reflective or impulsive activity? [J]. Journal of Research in Interactive Marketing, 2019, 13(3): 331-350.
[40] Glass G V. Primary, secondary, and meta-analysis of research [J]. Educational Researcher, 1976, 5(10): 3-8.
[41] Kim J. Analysis of health consumers' behavior using self-tracker for activity, sleep, and diet [J]. Telemedicine and E-Health, 2014, 20(6): 552-558.
[42] Walker R J, Gebregziabher M, Martin-Harris B, et al. Quantifying direct effects of social determinants of health on glycemic control in adults with type 2 diabetes [J]. Diabetes Technology & Therapeutics, 2015, 17(2): 80-87.
[43] Ivanov A, Sharman R, Rao H R. Exploring factors impacting sharing health-tracking records [J]. Health Policy and Technology, 2015, 4(3): 263-276.
[44] Park E, Kim K J, Kwon S J. Understanding the emergence of wearable devices as next-generation tools for health communication [J]. Information Technology & People, 2016, 29(4): 717-732.
[45] Liu D, Guo X. Can trust and social benefit really help? Empirical examination of purchase intentions for wearable devices [J]. Information Development, 2017, 33(1): 43-56.
[46] Zhang Y D, Li D J. Research on barrier factors and influence mechanisms of consumer participation in quantified self [J]. Chinese Journal of Management, 2018, 15(1): 74-83.
[47] Hassan L, Dias A, Hamari J. How motivational feedback increases user's benefits and continued use: a study on gamification, quantified-self and social networking [J]. International Journal of Information Management, 2019, 46: 151-162.
[48] Cho W C, Lee K Y, Yang S B. What makes you feel attached to smartwatches? The Stimulus-Organism-Response(S-O-R) perspectives [J]. Information Technology & People, 2019, 32(2): 319-343.
[49] Chuah H W. You inspire me and make my life better: investigating a multiple sequential mediation model of smartwatch continuance intention [J/OL]. Telematics and Informatics, 2019, 43: 101245. [2025-01-04]. https://doi.org/10.1016/j.tele.2019.101245.
[50] Chen Z F. The influence of online social support in quantified self apps on consumer purchase intention [D]. Hangzhou: Zhejiang Gongshang University, 2020.
[51] Zhang Y, Wang X, Li Z Y. Research on the influence of quantified self on continuous use intention of wearable devices [J]. Sports Research and Education, 2021, 36(3): 90-96.
[52] Fan J, Wu L P. Research on the influence of online social support on chronic disease patients' continuous participation intention in quantified self [J]. Chinese Journal of Management, 2021, 18(4): 597-603.
[53] Li B K, Lu W J. Research on users' continuous participation intention in quantified self based on technology acceptance model [J]. Information Research, 2021(2): 45-51.
[54] Zhang J F. Research on the utility of quantified self and its influence on continuous use intention of exercise quantification tool users [D]. Chengdu: Southwestern University of Finance and Economics, 2021.
[55] Sun M, Jiang L C. Linking social features of fitness apps with physical activity among Chinese users: evidence from self-reported and self-tracked behavioral data [J/OL]. Information Processing & Management, 2022, 59(6): 103096. [2025-01-04]. https://doi.org/10.1016/j.ipm.2022.103096.
[56] Pilgrim K, Bohnet-Joschko S. Donating health data to research: influential characteristics of individuals engaging in self-tracking [J]. International Journal of Environmental Research and Public Health, 2022, 19(15): 9454.
[57] Xu X T, Zhu Q H. Research on characteristics and patterns of initial quantified self participants from personal health management perspective [J]. Library and Information, 2022(5): 51-62.
[58] Xu X T, Zhu Q H, Yang M Q, et al. Research on continuous participation motivation of quantified self for personal health information management [J]. Journal of the China Society for Scientific and Technical Information, 2022, 41(3): 229-243.
[59] Liao F. Research on the influence of technology affordance on continuous use intention of exercise quantified self apps [D]. Hangzhou: Zhejiang Gongshang University, 2022.
[60] Xiao Y C. Research on the formation of emotional attachment and continuous use intention of fitness app users based on SDT [D]. Dalian: Dalian University of Technology, 2022.
[61] Liu Q. Research on the influence of gamification elements on users' continuous participation intention in quantified self [D]. Hangzhou: Zhejiang Gongshang University, 2022.
[62] Körner R, Schütz A. Examining the links between self-tracking and perfectionism dimensions [J]. Current Issues in Personality Psychology, 2023, 11(4): 319-325.
[63] Xu X T, Han W T, Zhu Q H. User participation utility of quantified self and its internal action path [J]. Library Tribune, 2023, 43(7): 132-142.
[64] Bianchi C, Tuzovic S, Kuppelwieser V G. Investigating the drivers of wearable technology adoption for healthcare in South America [J]. Information Technology & People, 2023, 36(2): 916-939.
[65] Ke Y Q. Research on influencing factors of continuous participation intention of exercise app users in quantified self [D]. Hefei: Anhui University, 2023.
[66] Yu Y. Research on continuous use intention of exercise quantified self apps based on persuasive system design model [D]. Hangzhou: Zhejiang Gongshang University, 2023.
[67] Yu K C. Research on influencing factors of smart wristband users' quantified self behavior from human-computer communication perspective—based on extended privacy calculus model [D]. Shanghai: East China Normal University, 2023.
[68] Choi J R, Kim S. Predicting individuals' privacy protection and self-tracking behaviors in the context of smart health [J/OL]. Telematics and Informatics, 2024, 86: 102069. [2025-01-04]. https://doi.org/10.1016/j.tele.2023.102069.
[69] Liang R, Li Y, Wei W. Advancing elderly healthcare: adapting UTAUT2 with inhibitors for wearable device continuance [J/OL]. Library Hi Tech, 2024. [2025-01-04]. https://doi.org/10.1108/LHT-09-2023-0413.
[70] Cohen L, Manion L, Morrison K. Research methods in education [M]. 6th ed. London: Routledge, 2007.
[71] Song S J, Yan H, Wang H X. Meta-analysis of influencing factors on mobile library acceptance behavior among university users in China [J]. Library and Information Service, 2019, 63(10): 56-67.
[72] Zhang S W, Ju R P. Investigation on "Women's History Month" promotion activities in American public libraries from a gender perspective [J]. Library Science Research, 2024(10): 128-137.
[73] Lu W H, Tian F. Research on gender structure of scholars in China's humanities and social sciences [J]. Journal of Jiangsu University (Social Science Edition), 2023, 25(5): 113-124.
[74] Tan C H, Wang H X, Wang Y W, et al. Gender differences in research output of university faculty in library, information and archival science [J]. Library Tribune, 2023, 43(9): 65-75.
[75] Zhang J Y, Liu X M. Characteristic analysis and comparison of highly-cited scientists in chemistry between China and the United States [J]. Library and Information Service, 2024, 68(17): 112-121.
[76] Yang R X, Xu F, Shen J N, et al. Research on privacy disclosure behavior of social network users based on social penetration theory [J]. Library and Information Service, 2023, 67(6): 84-95.