# Cognitive and psychological externalities of the attention economy

The modern digital landscape operates on a business model fundamentally predicated on the extraction, retention, and monetization of human attention. This paradigm, widely known as the attention economy, transforms user engagement into a quantifiable commodity sold to advertisers. As digital platforms optimize their algorithmic architectures to maximize user retention, a complex array of cognitive, neurobiological, and psychological externalities has emerged. This report examines the structural mechanisms through which human consciousness is commodified, analyzing the measurable impacts of these systems on cognitive load, reward processing, and population-level psychological well-being.

## Financial Mechanics of Attention Capture

The commodification of attention is driven by specific financial metrics that dictate the design and algorithmic priorities of technology platforms. The primary metric of success in this ecosystem is Average Revenue Per User (ARPU), which reflects a company's efficiency in monetizing user traffic [cite: 1]. ARPU functions as a composite of product price (advertising rates) and sales volume (the number of active users combined with the duration of their engagement) [cite: 1, 2].

### Revenue Growth and Unit Economics

The financial scale of the attention economy is vast and increasingly concentrated. Meta, maintaining its status as a dominant digital entity, reported $32 billion in revenue in the second quarter of 2023, representing an 11% to 12% year-over-year growth largely driven by advertising efficiency [cite: 3, 4]. Between 2011 and 2025, Meta's global ARPU surged by 1,040.6%, climbing from a baseline of $5.00 to $57.03 globally, illustrating highly aggressive monetization despite market saturation [cite: 5]. 

ByteDance, the parent company of TikTok, generated $29 billion in revenue during the same quarter in 2023, demonstrating a 40% year-over-year growth that highlights the intensely competitive nature of attention capture [cite: 3]. TikTok's ad revenues specifically reached an estimated $13.2 billion in 2023 [cite: 3]. This revenue generation is closely monitored by venture capital and market analysts through strict unit economics, particularly the Lifetime Value to Customer Acquisition Cost ratio (LTV:CAC), which typically targets a 3:1 or higher benchmark to indicate sustainable scaling [cite: 6].

### Stickiness and App Store Optimization

Because user growth eventually approaches the demographic limits of the global population, platforms must continually extract more engagement time from existing users to sustain revenue growth. This aligns corporate financial success strictly with maximizing the time users spend in-app [cite: 2, 6]. To measure this, platforms track the ratio of Daily Active Users to Monthly Active Users (DAU/MAU), referred to as the application's "stickiness" [cite: 6]. While e-commerce platforms average a stickiness ratio of 10%, social applications frequently target and exceed 50%, requiring the average user to log in at least every other day [cite: 6].

This demand for continuous engagement has fundamentally altered app distribution mechanics. App Store Optimization (ASO) algorithms on both Google Play and the Apple App Store have shifted their ranking weights away from raw download counts toward post-install engagement metrics [cite: 7]. Applications demonstrating strong "Day 7 retention" (users returning a week after installation) and extended session lengths are algorithmically favored in search results, effectively punishing applications designed for brief, utilitarian interactions [cite: 7, 8].

### Global User Engagement Metrics

The success of these platforms in capturing human time is evident in global usage statistics. Research indicates that the average internet user spends approximately 2 hours and 23 minutes per day on social media, accounting for over one-third of total daily online activity [cite: 9, 10, 11, 12]. Over a full year, this equates to more than 260 trillion minutes of collective human attention allocated to social platforms [cite: 9].

[image delta #1, 0 bytes]



| Platform | Average Daily Time Spent per User (US/Global Estimates 2023-2025) | Average Monthly Time Spent |
| :--- | :--- | :--- |
| **TikTok** | ~58 to 69 minutes | 31.5 to 43.8 hours |
| **YouTube** | ~46 to 59 minutes | 28 to 29.3 hours |
| **Facebook (Meta)** | ~30 to 37 minutes | 16.4 to 19.7 hours |
| **X (formerly Twitter)** | ~11 to 35 minutes | 5.3 hours |

*Note: Estimates vary by region and measurement methodology, but the hierarchy of engagement heavily favors short-form video formats [cite: 9, 10, 11, 12, 13].*



## User Interface Design and Algorithmic Feedback

The capacity of digital platforms to sustain unprecedented levels of engagement relies on specific user interface (UX) choices and underlying algorithmic architectures designed to minimize operational friction and exploit psychological vulnerabilities.

### Frictionless Environments and Infinite Scrolling

Modern platforms utilize "frictionless" designs that remove natural stopping cues, making it cognitively easier for a user to continue consuming content than to disengage [cite: 14, 15]. The infinite scrolling mechanism ensures that the feed has no defined endpoint, preventing the brain from encountering a natural pause that might prompt self-reflection or task-switching [cite: 14, 16]. Actions such as liking, commenting, and sharing are mechanically integrated into simplified button gestures, transforming complex social interactions into habitual, low-effort behaviors [cite: 14].

Furthermore, the rise of the short-form video economy—exemplified by TikTok, YouTube Shorts, and Instagram Reels—leverages high completion rates. Videos under 90 seconds in duration retain viewers far more effectively than long-form content [cite: 15]. The autoplay functionality eliminates decision-making barriers entirely, maximizing engagement time by removing the friction of selecting the next piece of media [cite: 15].

### Personalization and Filter Bubbles

Behind the frictionless interface operates a highly tuned recommendation algorithm that acts as a "pleasure machine" [cite: 14]. These systems analyze historical data, dwell time, and demographic patterns to continuously predict and serve content that guarantees user retention [cite: 14, 15]. This creates an "internal loop" of continuous, unconscious browsing, distinguished from the "external loop" of initially opening the application [cite: 14]. 

The algorithm's strict prioritization of engagement over informational utility inevitably leads to the cultivation of "filter bubbles" or "information cocoons" [cite: 14]. Because outrage and extreme content have been shown to retain attention up to 43% longer than neutral material, algorithms systematically favor emotionally salient stimuli, narrowing the user's exposure to diverse perspectives [cite: 17, 18].

## Neurobiological Mechanisms of Addiction

The efficacy of algorithmic feeds in capturing attention is fundamentally rooted in the exploitation of human neurobiology, specifically the reward and motivation systems managed by dopaminergic pathways. 

### Variable Reward Schedules

The most profound mechanism driving digital engagement is the variable reward schedule [cite: 16, 17, 19]. Derived from behavioral reinforcement theory and identical to the mechanics utilized in casino slot machines, this psychological principle dictates that rewards delivered unpredictably produce a stronger and more persistent behavioral response than consistent rewards [cite: 16, 17]. 

When a user refreshes a feed or swipes to a new short-form video, they encounter an intermittent and unpredictable sequence of highly salient stimuli interspersed with uninteresting content [cite: 15, 16]. Neuroimaging studies, including fMRI scans of TikTok users, demonstrate that this unpredictability effectively hijacks the brain's reward pathway [cite: 15, 17]. The brain does not receive its primary dopaminergic spike from consuming the content itself, but rather from the anticipation and uncertainty of whether the subsequent swipe will yield a reward [cite: 16, 18].

### Re-evaluating the Dopamine Detox Narrative

In recent years, the concept of "dopamine addiction" and the subsequent "dopamine detox" have gained immense cultural traction. Popular narratives suggest that digital platforms trigger continuous "dopamine overdoses," necessitating periods of strict abstinence to "reset" neurochemical levels [cite: 20, 21, 22]. 

However, neurobiological consensus indicates that this framing is a fundamental misunderstanding of brain chemistry [cite: 20, 22, 23]. Dopamine is a naturally occurring neurotransmitter responsible for motivation, learning, and the reinforcement of behaviors; it is not a chemical of pure euphoria, nor is it intrinsically addictive in isolation [cite: 21, 22, 23]. Addiction is fundamentally mediated by dopamine, rendering the phrase "dopamine addiction" medically redundant [cite: 20]. 

Instead of an acute overdose, the continuous, rapid-fire stimulation provided by digital platforms leads the brain to downregulate its own dopamine production and receptor sensitivity to maintain homeostasis [cite: 21, 24]. This chronic deficit—not an excess—results in symptoms such as low motivation, anhedonia, and anxiety [cite: 21, 24]. Consequently, individuals engage in compulsive scrolling not to achieve a euphoric high, but merely to restore a sense of normalcy and alleviate the dysphoria of a downregulated reward system [cite: 21]. 

### Incentive Salience and Neural Adaptation

The neurobiological consensus aligns more accurately with the framework of "incentive salience," wherein cues associated with digital rewards (e.g., a notification chime, a red application badge) acquire a disproportionate, pathological motivational pull [cite: 24, 25, 26]. 

Chronic overuse alters neural circuitry involved in mood and impulse control. Specifically, it diminishes prefrontal cortex activity and connectivity, resulting in poorer decision-making and a diminished capacity to resist compulsive checking behaviors [cite: 24]. Conversely, structural and functional changes in limbic regions, such as the amygdala and nucleus accumbens, heighten cue reactivity and craving [cite: 17, 24]. For instance, dopamine activity within the ventromedial shell of the nucleus accumbens shifts dynamically from the reward itself to the predictive cues as behaviors become deeply habituated [cite: 23].

## Cognitive Load and Attentional Interference

Beyond the reward system, the attention economy imposes severe externalities on human cognitive architecture. The constant influx of algorithmically sorted information and the demand for continuous partial attention fundamentally alter how individuals process data, learn, and perform complex tasks.

### Cognitive Load Theory in Digital Environments

Cognitive Load Theory (CLT) posits that human working memory is strictly limited in capacity [cite: 27, 28, 29]. The theory categorizes cognitive load into three types: intrinsic (the inherent difficulty of a task), extraneous (the burden caused by poorly designed presentation or distractions), and germane (the mental resources dedicated to actual learning and schema formation) [cite: 27, 28].

[image delta #2, 0 bytes]

 

In digital environments, the integration of multiple media types, hyperlinked structures, and algorithmic interruptions massively inflates extraneous cognitive load [cite: 29, 30]. When task demands exceed working memory capacity, the individual experiences cognitive overload, leading to impaired performance, reduced reading comprehension, and mental fatigue [cite: 28, 29, 31]. 

Research indicates that heavy digital multitaskers—those habituated to rapidly switching between applications—take significantly longer to perform focused cognitive tasks [cite: 32, 33]. Their working memory becomes inefficient at filtering out irrelevant stimuli, leading to diminished executive control and compromised selective attention [cite: 32]. The architecture of social platforms intentionally cultivates this extraneous load; users must simultaneously navigate interfaces, process rapidly altering video formats, and manage the social implications of algorithmic feedback, leaving virtually no cognitive resources for deep processing [cite: 34, 35].



### Gaze Dynamics and Physiological Markers

Empirical evidence of this cognitive strain is directly observable in ocular behavior. Eye-tracking technology reveals that high cognitive load definitively alters visual search patterns and gaze dynamics [cite: 27, 36]. Studies examining real-world task performance demonstrate that when individuals are subjected to concurrent cognitive load (simulating digital multitasking), their eye-hand span is reduced, and they exhibit more frequent, erratic fixations on irrelevant objects [cite: 36]. 

Furthermore, microsaccade rates—small, involuntary eye movements occurring during fixations—fluctuate predictably based on mental workload. Increased task complexity and visual clutter on digital interfaces alter microsaccade amplitude and frequency, serving as a reliable physiological marker of the attentional interference generated by hyper-stimulating platforms [cite: 27].

### Metacognitive Deficits and the Verification Paradox

The behavioral consequence of cognitive overload is a measurable reduction in deep work and sustained attention [cite: 35, 37]. Metacognition—the capacity to monitor and regulate one's own cognitive processes—requires a state of reflective attention that the reactive nature of high-density digital environments systematically precludes [cite: 33, 35]. The constant interruption from push notifications forces the brain's default mode network to atrophy under the strain of continuous task-switching, with studies suggesting it takes over 20 minutes to regain deep focus after a digital interruption [cite: 17].

This dynamic extends to the reliance on generative AI systems. Recent longitudinal research highlights a "verification paradox" among digital workers and students: as individuals increasingly rely on AI to solve complex tasks to reduce cognitive load, their verification confidence declines [cite: 38]. The seamless provision of answers by algorithms engenders a "false self-efficacy"—users believe their competence has improved, while their actual cognitive mastery, critical thinking, and academic achievement degrade, resulting in a pronounced Dunning-Kruger effect [cite: 38, 39].

## Psychological and Clinical Outcomes

The neurological adaptations and cognitive limitations driven by the attention economy manifest in broad, population-level psychological externalities. The correlation between heavy digital media consumption and adverse mental health outcomes has prompted aggressive responses from global health authorities.

### Epidemiological Trends in Anxiety and Depression

Clinical data indicates steadily rising rates of anxiety and depression, particularly among younger demographics whose critical periods of neurodevelopment coincide with the ubiquity of smartphones. Data from the 2022 National Health Interview Survey indicated that approximately one in five U.S. adults experienced symptoms of anxiety or depression in the measured two-week period, with the highest prevalence among adults aged 18 to 29 [cite: 40]. 

Globally, the incidence of anxiety disorders among individuals aged 10 to 24 increased by 52% between 1990 and 2021 [cite: 41]. The Global Burden of Disease study identifies that during this period, Disability-Adjusted Life Years (DALYs) for anxiety disorders rose significantly, disproportionately affecting females in the 20–24 age bracket [cite: 41, 42].

| Health Metric (Global Age 10-24) | Baseline (1990) | Recent Measurement (2021) | Percentage Change |
| :--- | :--- | :--- | :--- |
| **Anxiety Disorder Incidence Rate** (per 100,000) | 708.02 | 883.10 | + 52.00% |
| **Prevalence Increase: 15-19 Years** | - | - | + 19.32% |
| **Prevalence Increase: 20-24 Years** | - | - | + 23.16% |

*Data source: Global Burden of Disease study, tracking trends in anxiety disorder incidence and prevalence [cite: 41].*

### Health Advisories and Warning Labels

The severity of these outcomes led U.S. Surgeon General Dr. Vivek Murthy to issue a major public advisory in 2023, followed by a 2024 call for congressional action to place tobacco-style warning labels on social media platforms [cite: 43, 44, 45]. The advisory cited specific research demonstrating that adolescents spending more than three hours a day on social media face double the risk of experiencing depression and anxiety symptoms [cite: 43, 46]. 

The Surgeon General's intervention underscored the unique vulnerabilities of the developing adolescent brain, which is neurologically primed to seek social feedback [cite: 44]. Algorithmic features such as follower counts, public metrics, and the promotion of toxic beauty standards severely impact self-esteem [cite: 17, 44]. Meta's own internal documents, leaked in recent years, explicitly acknowledged that Instagram worsened body image issues for roughly 32% of teen girls [cite: 17, 47]. Furthermore, chronic screen stress—where the brain interprets digital social threats as physical threats—triggers continuous stress hormone release, compromising sleep quality and exacerbating mood disorders [cite: 46, 48].

### Methodological Debates in Causality

Despite consensus among public health officials regarding the risks of excessive use, the academic community remains sharply divided on the exact causal relationship between the attention economy and the global mental health crisis. 

Social psychologist Jonathan Haidt argues in *The Anxious Generation* that the transition to a "phone-based childhood" in the early 2010s is the primary causal driver of adolescent mental illness [cite: 49, 50]. Haidt points to rapid declines in mental health that correspond precisely with the widespread adoption of smartphones, front-facing cameras, and algorithmically curated feeds, asserting that social deprivation and digital addiction act as direct causal agents [cite: 47, 49, 50].

However, this hypothesis has faced rigorous methodological critique from peer-reviewed researchers [cite: 49, 50, 51]. Critics argue that many analyses mistake correlation for causation [cite: 50, 52]. Independent meta-analyses frequently suggest that while a relationship exists, the effect sizes linking social media use to clinical depression are relatively small when confounding variables are controlled [cite: 47, 51]. 

The critiques center on the "blending" problem: many studies dilute results by combining heterogeneous technologies (grouping television with algorithmic social media) or blending clinical outcomes with general well-being metrics, which masks the specific harms of algorithmic feeds [cite: 47]. Furthermore, critics point to the directionality of causation, arguing that deteriorating mental health may drive increased screen time as a coping mechanism (reverse causality), rather than the screen time exclusively initiating the decline [cite: 47, 50]. Defending the causal model, proponents argue that high-quality longitudinal studies do show a forward relationship from heavy usage to subsequent depressive symptoms, suggesting the "precautionary principle" should guide policy rather than waiting for absolute epidemiological certainty [cite: 47].

## Media System Dependency and Algorithmic Reliance

To understand how populations become locked into these ecosystems despite the documented negative externalities, sociology and communication researchers utilize the Media System Dependency (MSD) framework [cite: 14, 53]. Originally developed in the 1970s, MSD posits that an individual's reliance on a medium is dictated by its capacity to fulfill specific goals related to understanding, orientation, and play [cite: 53, 54].

### Pervasive Ambiguity and Heightened Emotions

In the contemporary context, MSD has been updated to model "algorithmic dependence" [cite: 14]. In a digital era characterized by high uncertainty and rapid social change—termed "pervasive ambiguity"—users increasingly turn to algorithmically curated feeds to orient themselves socially and politically [cite: 54, 55]. The algorithms, optimized purely for engagement, exploit this dependency by serving emotionally salient, polarizing, and outrage-inducing content [cite: 18, 54].

This creates a highly damaging cognitive feedback loop. Ambiguity drives the user to the platform; the platform serves emotionally heightened content that captures attention; the heightened emotional state reduces cognitive processing depth, making the user increasingly susceptible to heuristic thinking and misinformation; and the resulting anxiety reinforces the need to continue seeking orientation through the app [cite: 54, 55]. Empirical studies measuring this dynamic demonstrate that algorithm dependence facilitates severe filter bubbles, which paradoxically limit exposure to diverse perspectives and reduce actual news knowledge, despite high media consumption rates [cite: 14].

## Global Contexts and Occupational Externalities

The cognitive and psychological externalities of the attention economy extend beyond consumer social media, profoundly impacting occupational environments and global health systems, particularly in low- and middle-income countries (LMICs).

### Digital Fatigue in Healthcare

The integration of digital health interventions across Latin America, Africa, and Southeast Asia illustrates the dual-edged nature of digital transformation [cite: 56, 57]. While telemedicine and digital tracking tools aim to optimize workflows and facilitate active remote monitoring for long-term conditions (such as rheumatoid arthritis or schizophrenia), they frequently induce severe digital fatigue [cite: 56, 58]. 

Systematic reviews of healthcare interventions in these regions indicate that persistent technical limitations, complex interfaces, and the sheer volume of data input dramatically inflate the cognitive load placed on medical providers [cite: 56]. Without user-centered design and adequate training, the intended efficiency gains are offset by technological burnout, compromising the quality of healthcare delivery in already resource-constrained settings [cite: 56]. Cost-effectiveness analyses of remote monitoring apps confirm that while they can facilitate early intervention, they require careful calibration to avoid overwhelming both patients and providers with constant symptom-reporting demands [cite: 58, 59].

### Workforce Disruption and LLM Integration

Similarly, the rapid integration of Large Language Models (LLMs) and generative AI into the global workforce represents a structural shift in occupational cognitive load. Current data suggests that AI primarily reshapes job functions through augmentation rather than outright displacement, drastically altering entry-level cognitive roles [cite: 60, 61]. While AI adoption has been shown to boost immediate productivity and close skills gaps for certain tasks, researchers warn that the over-reliance on agentic tools risks eroding "durable skills" such as critical thinking, emotional intelligence, and complex problem-solving [cite: 60, 61]. The challenge for future workforce architecture is maintaining human cognitive resilience in an AI-saturated environment.

## Structural Interventions and Recommender Ecosystems

Addressing the externalities of the attention economy requires structural changes to the digital environment. Researchers in Human-Computer Interaction (HCI) and Computer-Supported Cooperative Work (CSCW) are actively exploring alternative architectural models that prioritize user well-being and cognitive autonomy over raw engagement metrics.

### Designing User Friction

If frictionless design facilitates compulsive, unconscious use, the deliberate introduction of "user friction" serves as a primary regulatory mechanism [cite: 14]. User friction involves designing natural pause cues into interfaces, enforcing transparent feedback alerts regarding time spent, and providing proactive user controls over algorithmic curation [cite: 14]. By forcing the user to make an active, conscious choice to continue engaging, friction disrupts the automaticity of the internal loop and helps restore metacognitive awareness [cite: 14]. 

However, implementing these interventions is challenging due to observer effects. Studies utilizing smartphone logging and Experience Sampling Methods (ESM) to track digital behavior note significant reactivity: users temporarily reduce their screen time when they know their usage is being monitored, complicating the assessment of long-term digital detox interventions [cite: 62]. 

### Decoupled Recommender Ecosystems and Pluralism

At a systemic level, researchers propose dismantling the monolithic nature of platform algorithms. Current social platforms operate as enclosed "walled gardens" where a single corporate entity dictates the singular recommendation logic for billions of users [cite: 63, 64, 65]. An emerging structural alternative is "algorithmic pluralism" or the use of "middleware" [cite: 63, 65]. 

Advanced simulation frameworks, such as SMORES (Simulation Model for Recommender EcoSystems), allow researchers to model digital environments where users possess "algorithmic choice" [cite: 63, 65, 66]. In a decoupled system, the content hosting infrastructure is separated from the recommendation engine [cite: 65]. Consumers can select from a marketplace of third-party algorithms tailored to different goals—for instance, choosing an algorithm optimized for educational diversity or balanced political discourse rather than one optimized for maximum dwell time and outrage [cite: 63, 65, 67]. 

Simulations utilizing large datasets (e.g., MovieLens 1M) demonstrate that decoupled recommender systems significantly improve utility and fairness for niche consumers who are poorly served by mainstream engagement algorithms, while also fostering a more equitable environment for independent content creators [cite: 63, 65, 66]. Implementing such systems requires robust legal frameworks for data portability—allowing users to securely transfer their profile and preference data between different algorithmic providers without losing their social graph [cite: 63, 64].

### Policy Sandboxing and Collective Auditing

Regulatory approaches are beginning to mandate transparency. Initiatives like the Digital Services Act (DSA) require online platforms to report on content moderation decisions, prompting HCI researchers to develop "policy sandboxing" frameworks to test governance impacts before full implementation [cite: 68]. Furthermore, systems like WeAudit are being deployed to allow end-users to collectively audit and report biases and harms in generative AI and recommender systems, decentralizing the oversight of algorithmic power [cite: 69, 70].

The transition from an attention-extractive business model to a pluralistic, human-centered digital ecosystem remains the central challenge of contemporary technology policy. By clearly defining the specific mechanisms of commodification—from the macroeconomics of ARPU to the neurobiology of incentive salience and the cognitive limits of working memory—stakeholders can design and mandate interventions that protect human cognitive autonomy and psychological health.

---

## Sources

1. [Bytedance Vs Meta: The Digital Arena Revenue Race](https://www.ocbridge.ai/bytedance-vs-meta-the-digital-arena-revenue-race/)
2. [Meta - Average Revenue per User (ARPU)](https://en.macromicro.me/charts/32031/fb-arpu)
3. [Social Media Stocks: One Metric Shows Meta’s Clear Leadership](https://www.forbes.com/sites/bethkindig/2024/01/11/social-media-stocks-one-metric-shows-metas-clear-leadership/)
5. [Meta Statistics 2026: Key User, Revenue & AI Growth Insights](https://sociallyin.com/meta-statistics/)
6. [Brain Spirals: What is Dopamine Addiction?](https://www.spiralsyllabus.com/blog/brain-spirals-what-is-dopamine-addiction)
10. [ScreenStrong Research](https://screenstrong.org/research/)
14. [WHO: Teens, screens and mental health](https://www.who.int/europe/news/item/25-09-2024-teens--screens-and-mental-health)
18. [Digital Well-being for Youth - Hogg Foundation](https://hogg.utexas.edu/digital-well-being-for-youth)
21. [CMU CSCW 2025 Contributions](https://hcii.cmu.edu/news/cmu-cscw-2025)
23. [CSCW 2024 Snapshot](https://medium.com/socialdynamics/cscw-2024-12432e4d12ba)
24. [Eye-Tracking Metrics to Measure Cognitive Load](https://pmc.ncbi.nlm.nih.gov/articles/PMC12668483/)
25. [Cognitive Load Theory in Translation](https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2025.1570929/full)
26. [Cognitive Load Theory Implications for Instructional Design](https://www.researchgate.net/publication/390000832_Cognitive_Load_Theory_Implications_for_Instructional_Design_in_Digital_Classrooms)
29. [User Friction in Infinite Scrolling Algorithmic Feeds](https://www.researchgate.net/publication/398444812_User_Friction_in_Infinite_Scrolling_Algorithmic_Feeds_Examining_the_Impact_of_App_Use_and_Algorithm_Dependence_on_News_Knowledge)
30. [The Anxious Generation Research Summary](https://www.anxiousgeneration.com/research/HaidtRausch_WHR_2026_Jan14.pdf)
31. [Recommender Ecosystem Design](https://arxiv.org/html/2508.03792v1)
33. [HuggingFace: Group-Level Feedback](https://huggingface.co/papers?q=group-level%20feedback)
35. [Multistakeholder Impacts of Profile Portability](https://arxiv.org/pdf/2604.21750)
36. [The Anxious Generation Debate](https://www.platformer.news/anxious-generation-jonathan-haidt-debate-critique/)
37. [Guardian Review: The Anxious Generation](https://www.theguardian.com/books/2024/apr/27/anxious-generation-jonathan-haidt)
38. [Critiques of Jonathan Haidt's The Anxious Generation](https://www.reddit.com/r/sociology/comments/1gtaclp/what_are_the_best_critiques_of_jonathan_haidts/)
39. [Book Review: The Anxious Generation](https://benchugg.com/writing/anxious_generation/)
40. [Increments Podcast: Social Media and Mental Health](https://www.incrementspodcast.com/82)
44. [Can the Dopamine Detox Trend Break My Digital Addiction?](https://news.cuanschutz.edu/news-stories/can-the-dopamine-detox-trend-break-my-digital-addiction)
45. [New Research Challenges Dopamine Detox Trend](https://scitechdaily.com/new-research-challenges-popular-dopamine-detox-trend/)
46. [Neurobiology of Digital Addiction](https://pmc.ncbi.nlm.nih.gov/articles/PMC12857582/)
47. [Dopamine Addiction: Fact or Clickbait?](https://recovered.org/blog/dopamine-addiction-fact-or-clickbait)
49. [Social Media Use and Cognitive Load](https://www.deanfrancispress.com/index.php/hc/article/view/1421)
50. [Digital Media and Cognitive Processing](https://www.ijfmr.com/papers/2025/2/41363.pdf)
52. [Social Media Integration and Cognitive Load](https://www.researchgate.net/publication/391434843_The_Impact_of_Social_Media_Integration_on_Cognitive_Load_and_Information_Processing_in_High_School_Classrooms_A_Literature_Review)
53. [Attentional Interference in Digital Reading](https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2025.1671214/full)
55. [Active Remote Monitoring Cost-Effectiveness](https://pmc.ncbi.nlm.nih.gov/articles/PMC12552658/)
57. [Smartphone-Based Rehabilitation Adherence](https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0000359)
59. [CDC: Symptoms of Anxiety and Depression Among Adults](https://www.cdc.gov/nchs/data/nhsr/nhsr213.pdf)
60. [Implicit Associations of Anxiety Trends](https://pmc.ncbi.nlm.nih.gov/articles/PMC12366797/)
61. [Global Burden of Anxiety Disorders](https://pmc.ncbi.nlm.nih.gov/articles/PMC11651023/)
62. [Global Incidence Trends of Anxiety Disorders](https://publichealth.jmir.org/2024/1/e49609/)
63. [Anxiety Symptom Scores and Loneliness](https://www.mdpi.com/2076-0760/13/11/623)
64. [Simulation Model for Recommender EcoSystems (SMORES)](https://arxiv.org/pdf/2604.21750)
65. [Multistakeholder Impacts in SMORES Framework](https://arxiv.org/html/2604.21750v1)
66. [SMORES Profile Portability and Fairness](https://arxiv.org/pdf/2509.22660)
67. [Alternative Recommender Ecosystem Designs](https://arxiv.org/pdf/2503.03606)
68. [Mechanism Design in Recommender Systems](https://anasbuhayh.github.io/)
69. [Media System Dependency in the Digital Age](https://www.mdpi.com/0718-1876/20/3/195)
70. [Algorithm Dependence and News Knowledge](https://www.researchgate.net/publication/398444812_User_Friction_in_Infinite_Scrolling_Algorithmic_Feeds_Examining_the_Impact_of_App_Use_and_Algorithm_Dependence_on_News_Knowledge)
71. [Heightened Emotions Social Media Dependency Model](https://academic.oup.com/anncom/advance-article/doi/10.1093/anncom/wlag005/8497640)
72. [Heightened Emotions Model Full Text](https://academic.oup.com/anncom/advance-article-pdf/doi/10.1093/anncom/wlag005/67112414/wlag005.pdf)
73. [GenAI Dependency and Self-Efficacy](https://pmc.ncbi.nlm.nih.gov/articles/PMC12561201/)
75. [High-Density Environments and Cognitive Development](https://www.scirp.org/journal/paperinformation?paperid=151339)
76. [Cognitive Load Affects Gaze Dynamics](https://www.researchgate.net/publication/389514771_Cognitive_load_affects_gaze_dynamics_during_real-world_tasks)
78. [Cognitive Load Theory Scope and Limitations](https://pubmed.ncbi.nlm.nih.gov/41594830/)
80. [Metrics That Matter to VCs: Mobile Apps](https://www.dogtownmedia.com/metrics-that-matter-to-vcs-building-analytics-dashboards-that-show-growth-potential-for-mobile-apps/)
81. [App Store Optimization (ASO) Dynamics](https://appradar.com/academy/what-is-app-store-optimization-aso)
82. [Global Running Apps Market ARPU](https://dataintelo.com/report/global-running-apps-market)
83. [SaaS Metrics and KPIs for Startups](https://onix-systems.com/blog/saas-metrics-and-kpis-for-startups)
84. [ASO Glossary and Mobile Metrics](https://asoworld.com/aso-glossary/)
85. [Reactivity to Smartphone Logging and Experience Sampling](https://www.researchgate.net/publication/397581738_Somebody's_Still_Watching_Me_Reactivity_to_Smartphone_Logging_and_Experience_Sampling)
86. [Reducing Gadget Use Among Children](https://journal.perbanas.id/index.php/pcs/article/download/1000/588/2741)
89. [Job Displacement and Workforce Adaptation](https://digitalcommons.lindenwood.edu/cgi/viewcontent.cgi?article=1767&context=faculty-research-papers)
90. [DataReportal: Time Spent on Social Media](https://datareportal.com/reports/digital-2024-deep-dive-the-time-we-spend-on-social-media)
91. [Average Time Spent on Social Media](https://electroiq.com/stats/average-time-spent-on-social-media/)
92. [UMaine: Social Media Statistics](https://umaine.edu/undiscoveredmaine/small-business/resources/marketing-for-small-business/social-media-tools/social-media-statistics-details/)
93. [Exploding Topics: Social Media Usage](https://explodingtopics.com/blog/social-media-usage)
94. [DemandSage: Average Time Spent on Social Media](https://www.demandsage.com/average-time-spent-on-social-media/)
95. [Neuroscience of Social Media Algorithms](https://richmondfunctionalmedicine.com/neuroscience-of-social-media/)
96. [Anxiety in the Digital Age](https://www.dsquaredworldwide.com/blogs/news/anxiety-in-the-digital-age-how-social-media-algorithms-are-rewiring-our-brains-and-what-to-do-about-it)
97. [Attention Economy and Algorithmic Reinforcement Theory](https://www.preprints.org/manuscript/202512.2551)
98. [The Psychology of Social Media Algorithms](https://intellitrongenesis.com/2025/10/10/the-psychology-of-social-media-algorithms/)
99. [The Short-Form Video Economy](https://www.grandviewresearch.com/blog/dopamine-loop-why-everything-now-reel)
100. [Digital Health Interventions in LMICs](https://www.researchgate.net/publication/395536096_Reducing_the_Burden_and_Enhancing_the_Effectiveness_of_Digital_Health_Interventions_in_Low-_Income_and_Low-and_Middle-Income_Countries_A_Systematic_Review_of_Literature)
101. [AI in Education: Southeast Asia Perspectives](https://www.tandfonline.com/doi/full/10.1080/2331186X.2025.2553835)
102. [Stanford HAI: AI Index Report 2025](https://hai.stanford.edu/assets/files/hai_ai_index_report_2025.pdf)
103. [Cognitive Load in Multimedia Learning Environments](https://www.researchgate.net/publication/334227969_Cognitive_load_in_multimedia_learning_environments_A_systematic_review)
106. [US Surgeon General Social Media Warning Label](https://yipinstitute.org/policy/social-media-warning-label-by-surgeon-general)
107. [Nebraska Med: Surgeon General Suggests Warning](https://www.nebraskamed.com/health/healthy-lifestyle/behavioral-health/more-social-less-media-surgeon-general-suggests-warning)
108. [APHA: Surgeon General Calls for Warning Labels](https://www.apha.org/publications/public-health-newswire/public-health-newswire/articles/2024/07/05/social-media-dangers)
109. [CU Anschutz: Social Media Warnings and Teen Mental Health](https://news.cuanschutz.edu/news-stories/surgeon-generals-call-for-warning-labels-on-social-media-underscores-concerns-for-teen-mental-health)
110. [Health Policy Ohio: Warning Labels on Social Media](https://www.healthpolicyohio.org/health-policy-news/2024/06/21/us-surgeon-general-calls-for-warning-labels-on-social-media)
111. [How Actions Create Preferences in AI](https://www.researchgate.net/publication/281549433_How_actions_create-not_just_reveal-preferences)
112. [Toward a Neurocounseling Paradigm in Addiction](https://www.researchgate.net/publication/392016460_Toward_a_Neurocounseling_Paradigm_The_Science_Behind_Step_One_in_Addiction_Recovery_Global_Journal_of_Medical_Research_A_Neurology_Nervous_System_Toward_a_Neurocounseling_Paradigm_The_Science_Behind_S)
113. [Expert Report on Social Media Addiction Neurobiology](https://archive.org/download/gov.uscourts.cand.401490/gov.uscourts.cand.401490.2536.4.pdf)

**Sources:**
1. [macromicro.me](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHywyOQ3W41zdH9miiZsMY0--wv-Q1dab2x-dBUZOK8d5_xBWwqsDREj5iNqAB10HC1AFUjE2JOtZjBTEmgc2wdv4azA8OaSnAMXqRkcz4FpK3KVWTB3TVXL69SXCknf0Thorw=)
2. [onix-systems.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEGR5CRmKKs0g38juRW7wvDTfAhpbTm-xuT1wKD_mtJqvkr61usIoOJ12imoaMjcykRL0E0a64-hX4bkhRIqt5O81Q6FiwN1IEC-Qlzu4psTvIQ1Lima1QrLWec1YKEhpEggLQgreAOlDIHQ1Gt5mmaJzyvCNxM)
3. [ocbridge.ai](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQETYuTgVS7LBlpQUCr3ufyod7B1rV_-3Hn29aA2OAoteTG-b2HM74OIFKHSJfrO6e_PpFmdwMXiXmISWkIkn7dyjxxjm2eDDu39tsYrAhL9gw-EXPJWwVlzqUQuxOfwYTa9x2jV_lB4LP9l_zNASLwZMECpT3mWpkjG2EXJZF1Y)
4. [forbes.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGSW-PHK3QvV5nnFuFdwKDbFw_6hyYL_VArzFiDanAXy3necHim7I9cXfBJ03tP4kUGgwnrVmNU_8wTKxRz10m-AYgaPvj8AU4AYJZW4h6trKbvtTQN_EXLQbSIsHXi3FtfgEDvEA87Jos6Q7fVS4hVSOCj38kPnFSNGbtSv-rjXubZHv5-pZtUYgYDyEdhWWhbHhvDAF5-EbfsKZhPoYw1C6j_qXk=)
5. [sociallyin.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQE1hHFMFLi8B3UcA5yPxdrSK_DauuOMQCMBcuV8fzwdl2HpC-QaAhQWyb3TlIjXwZkIObhTdRuMneSPGbYgJqGROvYQHigQ-LStWvcl6f0n4LwVN2PZ78tDopm0dEU=)
6. [dogtownmedia.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHh4Qg545ECnS9v_pV6uvRymncRfxL_wnDlssHFo3KYCVgmPKG7ngUX8DG5J2eHZSuzuOcfgwRdIqpZFMghV0vBeoPdYbmvKMJMga00ReDDDlVthuDy-eTN8kewldQqGfDB2zfeeTMdeCGyFu-7wmKTSA104k5xM_yM3QV394ToIE6j6Ce1H348Ya4WVz1KSJgdUVZyAxszAP2VZgbPHCzKaagpiXyd1YFMyLJyywzD-YbrJRphQUyM)
7. [appradar.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFqK7QrW8BqdMTkpPG89iiBm1QHPJk7_dkd4xeoJ7n2PwLLXlt6B0bAFTyBa1cndL9xcJlFDQCNrWHoWjfkH449pph4OMa6c6TJp8qn9_kpz66oNZHQu1GPzSS4D0Yn8w10kfsGVI59_pa_E9y0JjHt62jwebk=)
8. [asoworld.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEmd67N1GaAhZJYUc-F7v9_4iFrvFx97avMoPFdGzQ9K6MMAua9Afsc_rbwmL0LKMdvyqY9a15gVYpOJgKGe4Kg29TqfejuAw9PXWUXmDc4lkn2WRw_FcZc)
9. [datareportal.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFqVB_w-2NJe6WdOQpgskC1nz_ABu-AnvEglpfoNtwisJ8B4h3Zlewlq4GWM7sNEghlYLdl9_D3byv3E_yFmPI7C31TKgcQSEyJqYNeUgsHxY9Hb8eMB7nr_pCgbjS7kvi6qUSW_qgdUfwXVXHq3sEomKYkSyuc9glG6gVW4Ck87M9VEy7Wh5C2vLKjQc-VCw==)
10. [umaine.edu](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFSFT81T6N3sVCMoq-2baeZ_oAgjYAiY-Br60wbzKefIhTWaRSxn7bHJiuWQWckTzWkGwIa02jxyy5DcBH7sm3GNpXZvFAS7SkJjDMj2jwbb8p1LBQvlWvvwzPVPv7mHkZCuMvlpFOvzjVJtzrMgpK4Qy81mtCj3SmxfTcwMHCRmnJHZ-Nakir22bdl4FXhnDexyUcAV97yxnPOFP57O-XKTZMoZwevtCnzSVxas_cmgiYJgin5E1Gdjf26AfwBZjfo6U_L4g==)
11. [explodingtopics.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFuzioUTTPdh1lJd60Ydecj8qQFJoK03k7l-Xuj6P4xlF5Yhtx9LxABVNT08TBDIwN5xd7JET3t6fxGBcaBADm9H-Zr4f8re1rK-LCAoBM6dSnMgwCJgzE3Ct5pOpg7aVGs4Vy7xtsJie0=)
12. [demandsage.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGWbpaimEPFo1xWyGKGjzjn_SKAb1QyaeMQIUdihTln01qIDnOpoHaJ_-4YK5Kd-q0--QVlf7GDyup3hA31hTNRofJXRWOqmgpWCXZWDBdn5f4GB8noWj_sf-yg4i7jIMRLvxx3oyITEvFKmM4OPHFzRbRZnw==)
13. [electroiq.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFyxfi3f_OKoUXBVodmXgU-57UDx-sxaVy69h57XYX5JI6o1pNIKXzblIp8Q3M1LPrPhCafvucq-YDlTgO7p_F_yOCUVITNLFRW83Z3-kqlBfAdGUpi6nclrupeBjEWYBZAeAq6goTZOPixSqGcGWoJr1RAFvA=)
14. [Link](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFWC0NhdEdVVVCRuw0RtJUY8b2Xlzt-VLKimDhj85RQNzUUXY8vJFEXWSV70eRJicNNF6-wS1-dCiSAT8TSqYGHldJIctriRM3XV_i9-iXc0Z_1k_hBVjCAwE2WUWV319sfbUx_lB4C-t4V1ZYcTN0CcoGc6t01kxKAgv-_GgnDNoUAcsG7GMmL3LVajPsVUn_OyPoztAC4xCTCFcn_J3UqVvS3YDEzcTD7fGZLXkBushcG8y3FNvgUahRfyjUBgrmwXh8J1w5Q4qOLjcqk3yZ7ycLr641X7XjbAU6nJScJPb0Zx_D7TTjjz8Q=)
15. [grandviewresearch.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGaZVzpJBAyuPA66FMrhJwCQVquUN6-m8iWzbareKk2VZ6vm2wPAqfIrdwl9ZQVQSGRuE03p96NfHmJL8v83DzD8fg1de6RmhaDBCzhG_HUnvvEC580Gx1MnZvpFsxkZG3eHmUcRfSq2Gwx73Dwpi_lRKz9sNwzMZrz7yeiDU3C6Tpq)
16. [richmondfunctionalmedicine.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGY6dU3z7VQzLF0CIpWppkkSmUl7xoP-jG-_o5100zKw3FYGUvMEyuIhpUavwHjcda11rHfID3SGVTNwxYRr5ITh5JRKj9HyYuoOxAX7bTURL6zX1AK22WPbsGzeKKoaiiFwxUhJkU4Z7NZeas1CG0bs0qm_QJXcBSyWA==)
17. [dsquaredworldwide.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHmcsjWDjZkieeNc5VRHFfOJqz7kQKceLLLO7kQ2uplqnwzAgHNVjvnxtA0uqwQMCmSgXyfyXtxQSUSUzJDZ2lbM8wJzXY-a1tN51Z6BDbNJMP2U_4qEm6i1d0Uj9krEa5wCVyj9NHKnsGRjZmor-iPZL_ZcpZfd0rV20OPvCGKGDkViTgiUWg8ji4Ju1WfLUvOTgTQgkEWAPRuQwPibCPKd-WsSXPNJ_HWIKwlHnwRWmqJgHfTys1hAPRRuFuvrnYlCBK-_D06UGG1)
18. [intellitrongenesis.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGNxChq9cMEf8ZAaaPmJ4avIybiFnueXObni9yxgmo1hhTzI9qto3SJTpgJBkU1D4q8yzZOS8-2JggriP4jVvk99w4gjJVQfQ46Mc28TXlX4H84inAmjvDvI9qwfTz4PFp67W23Cb5lKKB9JQMy7uSyLPXNjyjqd6IobQfr47FsiZTzzbUMIQPrTgA=)
19. [preprints.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFDnLbkjqPizsMeYscTgFsi7NoYpUxFZLguYU34cPDK5MZDJVht9ctt-pMf5Zw2J1TVN9QKm7pCxervrwzCKaKmAxqd_LfTDBV5Eb80cRbwBj6Z1wverQyyQDheoAgXUpAVcRoS2q0=)
20. [spiralsyllabus.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHa65XDLR_CxwIgFDnDrrU0GCXoE9zp5oObdEdUXe8QKiy34m03k2ielfRTwvt4Szxzrjxhyx3e1MCksyOOlime5Wv1MXp4ySAmKJUTs5HsjoOrvmQYKUcyGlDeBQ7NEvx59cpIqRBnzef_NM5TKAaNht4_Wv8BGpqThGjOX3oy5BVS)
21. [cuanschutz.edu](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEPYzgcpogr06yQWxRbaVCMzn-kbnFf8S9ZZZEHbxwEP1YttSowwUDqDqLmR6qKEWNIWfQXqnQSRDXYYSTjed9kKyXHJEMCDCkWtTAIKk8XXBy24xJCaJceHGKPnKQUkoWMsdTy1aVyGxDUHn2rY89GyyjLmPVRhd3mG52wQV-d6afti28S8rU-4fbrQd5DPJP-1zdr0YM=)
22. [recovered.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFfTUfl-uxQzbQK4nPDStWdB-8LpdCun2aNftP4uxGpV_WR0xzCA4NNmTuofCN4bQ1wbnK8JBJqR4sjHWHuNK_JU8kQLIHowjVprtsXrxxLOsv5JxaEnlhC56Hn6N3r5ZGOlzy393EnUNdjakoSBR_caueiGKY=)
23. [scitechdaily.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEIm8ybFuph_GD8tdVL-MI-xyiW9R9SySUxFw5FkeGZ1cxrLmWORdG9zS30S10doD65VlHn7HXQypi81bhyPC5opL3WOxBaiA_sA0p3AzqYi7Ft5ttIYIAm6AZRyPQe6FOxkBm-jcw4-2Y-kUMlEUjqW-KFrFbMZXVxqjG_TgSNWim3wLA=)
24. [nih.gov](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFXifCM7Kpev-ZvniZE2o7FnKEFbVvgZBwg1ayWlv4iekr36U-Sfe4Z2XRAfqPKqsT5VIVYwy1Kx57CXmuO54KPqpIi12VffNDqpGpCXIOGhORZMqJx_A5TWMzwch-RI4mLTW9_VHVcmQ==)
25. [researchgate.net](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHFXFkpti2s_huWnvrRXqMJ3nmhYN3bHaxblAwPIkBjFbDA0Mf72TyCJVkYea2SaMqo8PyGMfMRmH8jkKrEvyyXpDoVtWWGYJcPjxP5e9J0deGe22DHO7eBYJbBYQbwKAALiKDeYbiDADG_v9mGuAW2T6ex0MsNFMCII17EMOZdM7tE7it7sxNh0eovO3xyTidSze2uFMYQHoPzgbtM1USO3-ImT1P9rKpzMxR0_WMlyfeR-WMDBmr99oD7f4NNMKVY-CguCNQSHTprBrjJjvsdOdfPdiXca-eJd8rGHjsc-kMUN0DtxGbMQsOfC-1_kPkIPn7pvJPankAsmiAJ-zhjfSMjerPguoYLLDODNbiQBwM5uYbcDRYO6GCpbRd_w6Sc6TfNPWyLmHA--iddfP9PYTA=)
26. [archive.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFHeRLEHbHjAHzw8Yqw8WQXcgBd1wH0ZQcOLhThlb_O5GoqpYvVtahBvDdFUCk9cJ-7ba1Te4zoClMpkI-PFsTeYqIEsHOcXWiMPmAQzjoDMKCV3sHotT6kOfTA0I9NUTfe1xwxcmN43qIUm3TTP-AEXyR4oxNDFMYUumuFcbWfBoTUlzydDvFBGJmSvC07IQ==)
27. [nih.gov](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQH5k196ywp35u3xpHWkTLR-_ZnRlJEohyaOOygve5fiK0co3PdKFWwjYMikmM9pWOfXCc4zPqxcY1C0pBjsGIEg_BwYL6aYueQ4a09v35ed5GQ38c6QPr7Tnbcq_uThUhldEcQJAQMZ1g==)
28. [frontiersin.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGJYZf0ktQYnmfFZcfSGJrblqIx6on8TQL8H26jBoj0jXO2UTyUORhvspw-0jcWv5oOO6ZJkv2eSqyr2IhDyx8-7ZD6Y6rx9RluTS5B8FT60pD_CWcByt5N2jn6OgSd5HFkZfK6oqrKHsSeHZGNyHLgBhlFc7cp7KwP9Qb3kPYm_AzUe8o3fKZktWU3KJVR)
29. [researchgate.net](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEAc_9ckBoiovf1hm9hOVV2CjylCU7SXkIRj2d5NOiwLs-8jqI1jj8bJbXT-v7YULXgAa-QE0NiiH-G4RCxxpgKUW05aYiuek7ZmaPLxQR-69Va3WfuSIVYShanL55oOJLWtpi-e02dH4oE45A1MmcXUVzWlFtH7Vqg4CMKBta096pgxIaCTQP2-VujawtagHMwH8156w2ZlJHJg-TrgNwzD6Ptsm486Cp0XGSeaTQukMaVIwAtak13g5OZ)
30. [researchgate.net](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHgGlD4_dayy5oV4ZkYXvJ-Jt51K1RB5T4Me-sAOYlHvQspvuTvFi5xTQCJTDBam6W6bh9QFPLo2q2RWJwG3gRyjBO6SRy0Xb2-UdQc0iCTaGdlsr3o9TBklqcNygnQhCgg9FWyQurfPnuipUgqalSGaXpcMLJ6KaygdQg_kyDq9OlclUneDgnPgUf1xBeV6mCdFmMrwo4sSH36OayYvubTicvLC29UqfwDIQxByUUKX0bKYmKCJ6DLckgpMfRawkmh13xhPxI_g32FHINl10ix8pH61cwG5COSd0oHdfVKS8NjGHtjvc3OOoER)
31. [frontiersin.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHNzolc5hmjCZ8YYBWVDfu87ukQ9wmpR2bcLi193mpdoDCU0TF_odzk4vQw4A98dTf_W9TcLK6dqOnms7KvYmA1RJS9d8hlNnQo6kNCiUqIdf0xouzaexI_QQqs1_3MqewhnawnkKnJ76eIqI1CCLJpNEI6Ul28FUMUmklSN9XvsxjP9FNicSKlDahv04_d)
32. [ijfmr.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQE2Ff-FzUKIS0s7JXVpKDycTwkRbR5ysuZUmHJku0cpYlMCLzw2kbnV-tpXx2xoe0oLKW0XQX1qR0w-5cDgNovBmTBJepjRa-eSqhMDZcBMtzBdU4qrUZiIddDQsdH9jWOWJIc=)
33. [nih.gov](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFpEZQynytB3_LT3jg6eTciSyK4CBqD-eBJ39IyzeJZ2c_KYsNwONt3oTpO6n-rvjLCUwfTHjxRz9Q_qXtWRVXF1ALHBqjH65tjYhjDmNPUAtkVd6BmIyGpyKq7TWYVYQ==)
34. [deanfrancispress.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQF0m6fzXfXLMLyDZdx2Pb7Xoaww0JR3jhQR56EYMmVMeBbL6mjogRSLUhbNsph96CG55cVg09o7RyURPpCRMRxaLxlItVZhf9cuJxCtd4Zd8KpwhlxnjXdcAYgxwvzQDIAQyF3yTJpIzDFjy39vE3XVO4Pm65I=)
35. [scirp.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEo7vuFSHo3jjwucH6_LQ2RmZk31m2Zgysv9kBcB1_vbjnAjroaEdFft-0TZ55ZjDsoLPmE1nCvSd_v8-UeiVK1rm0zwliGEo1WPjNZimR4zRvMKEGKSRdEusCjG_q4ZtVyseMorod846wSPGTqT2sC0Q7l)
36. [researchgate.net](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFvgB88X3Pqh0yt_bdWIUJcctyGbt2e98NlORBxCV5FnGnCie6IjvtEVxX7DiGoSeZ7GS9QUblZmSkRLiwB1xThEJ34wC7Sex8scqlROHW2wb_ZwlceVhKCGTAe6u24Qes-iw0wqYiqjTMDIxQNgbeEp7LBcmDkIR1ieRXxBJ0NlYQ2G-MlXR7sD7LlJB2AetzkPt_P8tuE4XsfkCPBf4lqZQXC87M=)
37. [perbanas.id](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQECYLGurwEtYyMyV5wrkntcHUKO3CuZ375RPqHhIHGg8ikSO3StO_WAoa-C6vhSoT7jBFiVtLmNH_SC32yEbjt94f5yIkUTkeQAgMRf78GQAQqcoQlF2l9cldl3jiXgT04UlkpqNiM5HsbqoQ7KmT_owCc3YzbyJPShuDKM47c=)
38. [researchgate.net](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGJmalZSYQb1I3Frsk-5E_H28l498TFu_PBxkSKaEXxQUyvnrPzwYBZR9nAL45aJD6iMJR3p1TJcwq8Y5TDjEmNeAkRYR-OerOIXDZVBVbls3czQnJB_XrrJjX27FA5z7h2m4OrsDSlU2jJPJ1pVtWDNXBgUA0f_IuSVs3uoaudzvzmX2ZR9T_w37uMJ6p6F50Kkyg7JpQS)
39. [nih.gov](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEo_HFLjQkjVKIn1r4gI1TwNhEuldygugsy02ZjStYPVTkgP-Sl9uMp4veoHvVVvTS2N2KPq55jhPWFtz9p8h7t3gX5Lf2_l_NQuomBMzs_CyMdLBbOHCpsOQZISECidGpd-N5JFwIy_A==)
40. [cdc.gov](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHRaMsYAW1DVj6VHvHU3vjk9kNwEk5rQXIcQlTck_wtuTZJJQIeZkXd_neMcxJf-NvY08Dkc7LdntO_bR0xuKshsrztxhvFuQgJhvFZWWunEgaESvz-e9uAa6Q76nEuyKa2xQdF)
41. [nih.gov](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGnWbrGP6f_2kUtZhRQnusL-gJnMsyRmypSc4h8_7XxhqlhAqq7cWiGfBUdZChOryfpeIUnog1OMSbnn4P01HDBpBADZzPo6o_1PS9j5rU3xXLlY4n9OPuxFuzR33nTy9W1zUihhMQzGg==)
42. [jmir.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHhkk-mcBlblK1dP4hquVh2J3JsMwlg81xPngTe69Rb3Z6qd3yX7e4kqn9ni5ST-65JnyLwF-lRVLxYYSBZO9Mq2lP6E4BOhIa2wZs3k1lE-Fe-Zd5r7ent_NQyZLAoX8dXnQ==)
43. [yipinstitute.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEq2YMpnnmPTk5rWpONN17wYM9q2msXRg3JUq5Z68chrKvc2nga_b7Vtch7Vi8efKND6ZvsV-pUTTf51hpmmvFicTU4-Eq0eoLNJHzbKqz9AO0ltmtAnOg8Cdn-mzRB3SL8QNm9s3vXtMoHwZsNeLORL-DJZtgzsp5ovNuiczpOiAzBMA==)
44. [nebraskamed.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGQcSK_cOvamuRX9LbNRWvxgY0HOX4brTJqiqslyBFJZcl8i27kKGhRfD-dFIMmXQCPWtXwOHbphbaQX5kPlvS1KezOIZisc9f48GytC8W6LYqK0_RaK61p8CIQvLnAF9eODCnWOzrAHoyMfuw4EwTX-ZgolI-57k130iohpjmfSJFnJUxYPkSKj-bVsQb3OxN_NiuaEpwRrDRpl3CgxU1vn8hm1s2IXZGZsTVW9nv0wWpn1CY=)
45. [apha.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFc1WlfuXPEvGKYNJVOETxRjS3RSA2BMp3GohNMk7lcdcEvUJJLUjXzC0bzpaqky41Dkpl5eVy9E4YipAqcGoTXJxReVegsH6qtbFqk6Q_FvxQ-g8Sf_sa7lXOwy-YTGHqv3RLKkOy48PDYAMvV7GQVXtmo1nd7pwDciOS7CEDgfggLGxpuaE2JOl_jC3pxYs1an4sHwh2OF9CYx34iQLTTODRvXuUlvBXhMCrZs3M=)
46. [cuanschutz.edu](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHqccaRLuHGj12cGj3tce7tfigQXLtDUWWiNw0yvCergwkA3r3vc8EvPxW3DYg-DzyXxQ87x7U6r8Y0OnCDZuZDlgj99OMYmVtHUVrcetk5dUG2rpuFXO9bJ6mHxoGwuWo9z7_EItp_26DwvDjM2KEgaV52jLJbWpRoKg2ImxdTdWgBRpi7wch6Ufgs-9D0uX2I04OmLbd-8k3a5jwC7vlR-BRM0CDMZypMcqbMxHXAc93nVCSAGVIExuEGo55cEeOMiq4Z)
47. [Link](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEXLOy_vk_QWMUjV_dv3o4BiQELl2BNKjKKdFwO2FRd4DAYTEL33C2wSazAqAz8dkjPPKgOxmQCQ0nVn7l-l_fFtoyZSadKsB3oZqd1gNYOZslBdn85NiABCI6mPBIoBRm8B3sFdIIayIHhJCCaWdyokh3T8XT4eQBvosC5xHI2)
48. [screenstrong.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQE35xIuBXx4BpNmT2sKEVNWexO3QoIMs6cmAuD3wlX3Bx7nTiDBCXcCHCHRvpm48vSWjxQdM2Pj3du9KI-ofXIOaGdpd50s5BwQBbmXmIn-_5EcIJ8x5a59)
49. [platformer.news](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFSa_4fjbHb4TExllBjdIRjldLEHes9nvAV2AeGZaWSA3_KokvQGrsReSHKgwItVjwA6RRjrMkscwpnd_zaUkeoFURf3GnSaMGU0uf7A9W4AX8vlqmi6POVoLFv0pbGxV0CkUDkY9UsWMNRpOl-0xc6eoqz6nUEUWgqiZGwnNMul05ztFk=)
50. [theguardian.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFDuN51vR5iYdyuekiDl20SrczsEUrw1wRPngdllMgfCu-WOHNdXCjZpc1mJiBC4bPQmRBnLnWuS8Mg13qC1gy8-q3UJqOoPEqv3XOAsJPKHpI8VTkajRX-eJeQqR9MLhDbOZ1ch8Uu_52jj8bUcVlU8Z0ac_JtPyVakFZbEk_C3AH5J661)
51. [benchugg.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHftumO7bj72UoNnct5jtyPNRE_YuihAQd06P7RyGCoUVIxkMXEbMp20suzlhUdWMowIzKcnRgULUORR8K6caDnIIomYRxtneElu6kSpbruMChbQKnCDumjfLRg_pT2OdDtyc8PIOw=)
52. [incrementspodcast.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGMddekEr6reJzAXPRmn_WPASb31wQ99vFBvqZa3RJVTTuzOPncZ02JBzDP5bwuGircixVjVgg3myROiLobsqjSHLdrfq3vzQOFfg0lmmN4tp4Rj7k9cR4KvMU=)
53. [mdpi.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQE6gZBTDoELlVpHFQOMXm5lZgcJS8wvyJNgYjJ5gyjHhfH7-zyAdv0hcC0Y5e3CpTG9snq0-sau3E6kBG1Q3_jIlv0dElXswOsP0ZlS5pFl89vKSEKjQZoFDLCYMj0=)
54. [oup.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEhrhNw8izMF0Ypoxt0KaapMSTq8_BfUEUIqW_0TnhLmFC1IQN3ole7YsKceMQVwaNQX93Ukwuim_O-WwjNsy6MpmwgTmDhwtb7164nF8sk8gWXrAaD7Iy9fE5tzUD4mNlftEZl8E3moH09OswLxNZyBdY8M7luWzFHTfnP4UwzbuxGfGuyfCs0)
55. [oup.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEQ4XX54v8LaL_Jq1nx5eldbeeeUcZmXWPWuK50VjSAQqc6MguKI36vvpeAs_J9tGd5pPVP04X-tcW0rUGrNdp5wtjTryoEmwMWFRGNGa7Ei30p-moxTIuHWzzbLfZVTfPrHudLjf9Jm3KSLfi6PcukW90EweNnR181VZk5cUEG96TvlxeTcGIrsUmTLI-HCEoQ5qJUkjSzyEg=)
56. [researchgate.net](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGrtCQP1XPK-mmkU8Grdv1uV1Xj_iugzPErmav7zTg8DAwwKX9xmIBMy4FPjceqwg0rFZfoZxCo7Q0I8S170TX48gctparkJKNI0C79i8Vpm3u8ogmuTnjo-6XiwPKoRrHKzi4mNUIGtHVTPISLi2EVR5k0x-HbN71kywJ3JnaUuBTJxhgV8MHa6seFjQS4UVGf_Onckz-9K8g27yQ70Y7KKzqNaIeywSSQ_xgvA9lDwamhYb7nn3SNkOO7RpQrevDRoWkfknbnuM9hC5XeLbRl3AGIL6ViupY42P6NhS8eOanNJN9TVTcAd_RBYrG2odP4Dz2GEGNN9uSGSBdzkxVpga11FQwNYRRK3oPb7GZp6OYI)
57. [tandfonline.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFRceTI_9SyQ2jFT4jU2f3XcxkKQQHN56IIZmw_LkF02LPxYYb9_7Q9biXx5wqKPSTN2UrWnTRPX6u4UlNQA0h_JzzX0mIgpi5bj3Gm9rQorLqY8blfVTP_T_UKtyH4ix53NP7bTOapw_agqVJMEaol2ZTYUQwzYzA=)
58. [nih.gov](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGTJU_CvNDsbuRUlxS3iq2YB_eFMCkORE5uhfyLRfCC-QGGxZSm7gNC-Y1-NfV5QMYxVT3yd9QD4rt8hVx9PDQHW7lcAAZS-qXiuApS9p4STfBevKpPHKkf369H0b7KYKAV9O4TZikWIA==)
59. [plos.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQFDWMWlJFph0fZPtJzVRZSF6kWPASxvM3IAGFzzDmTC33nEqY2Wxmu_-cVwwQFGeMW8IK0lgNpvYbd9cB1BB8YHLgul3EHKt7uLztwBpXM3x6TGdG5UFAVjpY7tA2JiT8LIULbA8KTGmx2qHjvvNj_CReuqBlDj-qeNFIIb9rAwzYTZ3FTc)
60. [lindenwood.edu](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGS8oCGXHZAex0rVsOB6btCmHAx2uBRiz4-m1M1U27J1KDlH4jfJ41b4XIguhD8A6eABXOmNUXnl8k8RT5I4e7_QnvfGHU0alUJtxDYn7dEPTFAwp6fhoUpaQBKb3zId6ksVH78JwpSH98sfqay-JQXcA4rsH8z5eUgFQq1RXVJtE-IUG4T58qVHSXB2dHITJkckgx4FtL_JdXuExU=)
61. [stanford.edu](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEwkYb17R7SmUmMP2ALcjqZAcMw6uyxifmfxYXXA-Z0pgXLmlIqlwE1c4trtvMOpH0DJi419quNNzTLmj03MeSO05TZ8PildfBaBEu_DOwL1gzOEkAvCwmPzTTVPaUxnOfAtdnbSgwcNsK3AlMukJE6guReIMe4zho=)
62. [researchgate.net](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQEk6jIc2-I2XHNCEYWPPlbI7DKs7z_cGicUgL7ZqjHbI2g1KoUfvwItC-4h5wRhms13yu6N8tz9RqdEvbCauC16Y2drbgEVjDe-FHyeHbYo0pDXK1xzu2nCaHyBIEvHHdUbMZfKQwG-oUdFtv8YBTFsHCGwrOQRAl6ImsbEcop50j72Tk73t2ulaThbUfKoZugxo9WicCX2lMSHwiz1akVP1w8hdL-yW8ZBd3V9E9aDXPInRAA7ccxzKLvMmqyhMg==)
63. [arxiv.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHRZGfvGKASZuwC18eNDkKiFawP2aXijvtTdmZO972BCWO3i5OdVNE4bx9VbvhvIwG-vkJW7iHVKNQEqVkFL7SQPHV8haIt5Ic302ttoxmac15nq_7vug==)
64. [arxiv.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQF5nqXwmRWTejW28wdzFHFYbWgpVw5-A3ppJ7cALf8SB4pZhPD0w7g1yFhLVUHxwuezwsFA0iFWqelGQ1a7U3-KeZ1oXaKau1Mp8XkyN_pyoxtddCe9stOmmA==)
65. [arxiv.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQESy_KGosZTl7nriO8YzakIbY7Pc99tSxp7e7wpL5X9ojOixFLDjv5Wzh-lUNNMQORTRJLxQB79C9nvtElhkrUOaEyULeNUmLGzcPpn8qQ7EaXkhPGuOQ==)
66. [arxiv.org](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQH2Y9BGTWo54POxaAUeLLo9BpaoGrmdJaigIV-2afMc2Yv7OaviYjjbHs1ORn715tNUDQUu4BG4ZltlfMhaVodvGiZht7n9zOaaaL5I9-XHAnajrod3VA==)
67. [github.io](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHvNTw0FNYZ41JtJCXLSt51qPsaRI6lLtqPtGvfWY5s6nu_Rm1xQ9_wZICcRmAMYlcuX0SZuhryxSQo6KXkcSF8xvx7mHasfNzk2oxnFLkCRYi_CA==)
68. [medium.com](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHSWnYHiZwP8Le5H9V9P83vfz1caQ5CUDkTiEoFlbhznmLWboOLV08dVG7ssT4fV8l3GiNLF7w2e2ogeuy8-KDhKCnZj_uBmvFiVRNUcehWXMTHjWRQhzikBOWpLIhdMNKEq3lk2ln2C3-hV_p2Ng==)
69. [cmu.edu](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQHY-mPc8ktrOLuim8EtYZ2YBErcEL9_YpojBJY8-uJEHs2_XxompwfxK0K3ZAaHpzT5UNgD8YaEmVJqReq5RR-oHftlXNTyK-L0HpHDL3NrsI-beGQKNYPhKGFQnAg=)
70. [huggingface.co](https://vertexaisearch.cloud.google.com/grounding-api-redirect/AUZIYQGiKwmdjGww83t-yRP0lF5iC6ONsbyBzfVGol9rOJa5IMncIqzA84NvpoyCYeEJJRKXfab_-J8XaiwG5Px0MvKyQpmu9olvSbMRQp0eYvpCzreYbLLjsVoztvf5s01Q5mbw8V8sWMUOIO9QSPI=)