In less than a generation, advances in digital media technology have transformed childhood. Together, children’s active consumption and passive exposure to digital media have gone from occasional to frequent to nearly constant to environmental. The breadth of this phenomenon has made comparisons among children with differing amounts of such media exposure, to assess media-related effects, practically impossible. In the foreseeable future, there will be few children who are not more or less continuously immersed in digital content throughout their waking hours.
Such social transformation challenges everyone interested in children’s health and wellbeing. New paradigms, more sophisticated studies, and more powerful tools and methodologies are needed to identify actionable causes and effects, and to design and employ precise, timely, effective interventions. Only with such advances can youth media effects research help children to optimize their benefits from digital media and mitigate their risks.
The Second Digital Media and Developing Minds National Congress convened experts with a common concern for children, but with different philosophical inclinations, methodological approaches, and disciplinary backgrounds. This created an ideal setting for defining, and addressing, scholarly and clinical questions about youth media effects through a kaleidoscopic lens. Attendees enthusiastically embraced this opportunity. Their presentations, panel discussions, and informal interactions produced insights, and identified opportunities, regarding a number of crucially important topics. One of this event’s most significant outcomes is a bold, collaborative effort to develop a new Media Screening Toolkit. It is planned to include a suite of innovative instruments for screening, monitoring, and measuring media habits (in toddlers, children, and adolescents), and for determining if, and how, those behaviors present in clinical settings.
The Future of Childhood and Technology
The current public debate over the virtues and vices of digital media looks a lot like how prior generations argued over their own emerging technologies. To make our conversations more productive, we need a new paradigm that reflects what we know about today and reasonably can anticipate about tomorrow. Modern digital life is social, fluid, and dynamic. Reducing all of our questions about it down to two options – “good” and “bad” is unproductive. We now must ask what forms our digital technologies might take (for example, beyond screens), where and when might we use them, with whom, with how much (or how little) human or artificially intelligent control, and to what ends. We may be using computers to do math, but we also may be discovering, informing, evoking, feeling, sensing, or moralizing. The devices and systems with which we act may be engaged in some of the same adaptive processes.
Early Childhood – Minding the Gap: Research on Digital Media Exposure in Early Childhood; Laps and Apps: Early Childhood and Parenting
The 100 billion or so neurons with which we’re born are all we’ll ever have. By age two, a child’s brain already is about 80% as large as an adult’s. By age three, it already contains a thousand trillion internal connections. The specific extent and character of that initial burst of network building differs from child to child in accordance with their individual experiences. So, too, does the ongoing refinement of each child’s neural network as it progresses from secure attachment to self-regulation and into adulthood.
Although many fundamental physical attributes of the brain are fixed by age three or so, our brains never stop changing. Individual and social experiences substantially define the specific character and ongoing network development within a child’s brain as he or she progresses from secure attachment to self-regulation and into adulthood.
Children’s interactions with their parents, caregivers, and environments (and each of their associate digital device and media behaviors and associated content) mediate this process. Ambient media in and outside of the home; parents’ and other caregivers’ use of digital devices in children’s presence; and children’s own direct access to such devices (for purposes ranging from education to pacification) are such environmental mediators. Dr. Hirsh-Pasek noted, “Everything that I study tells me that the contingent human-to-human interaction is really, really, really important. It's the glue that holds us together as little humans. Anything that destroys that is probably not good for kids. And anything that can prompt us to do more of it is probably great.”
We have yet to determine fully how technology plays a disruptive, adaptive, and/or potentiating role.
The progressive encroachment of screen-based experiences deeper into early childhood makes this area of youth media effects research a high priority. The large knowledge gaps illustrate the need to make more use of existing imaging and monitoring tools in early childhood studies. The increasingly social character of digital media use also demands that we look not only at how media affects individual children, but also how they affect children’s social capabilities and relationships. Other early childhood research priorities include: (i) more lab-based, controlled studies in neuroscience and basic learning, (ii) more basic and translational research in naturalistic settings, and (iii) more inclusive recruitment of study subjects, to take into account a more representative range of socioeconomic, racial, ethnic, and ability attributes.
Digital Media, Mental Health, and Relationships
Digital devices and media that represent change to older people are just status quo for younger ones. Some adults may still think that abstinence is an option, but that is inconceivable to children, teens, and young adults. Technology is too compelling a means for self-expression, instant gratification, relatively effortless social rewards, and timely information about current events. Although eliminating social media is impractical, some of its its dangers may be mitigated through more positive online experiences. Toward that end, researchers need to shift some of their attention from what young people are experiencing online to the offline effects of their online experiences.
Digital Media, Mental Health, and Wellbeing
A substantial majority of pre-teens are using social media. Their limited capacity for self-regulation, susceptibility to peer pressure, and tendency to seek autonomy from family networks all make them vulnerable in that environment.
There is evidence that teens’ and young adults’ mix of smart phone usage is changing. Email use is down, people have fewer friends on Facebook, and they’re spending less time there. Smartphone and general social media use and text messaging are up. Among high school students, having more social media apps use is associated with more frequent smartphone use, and more smartphone time is associated with lower GPAs, more anxiety when without a phone, more boredom and inattention in class, and less likelihood of having study strategies. Teen’s enthusiasm for face-to-face communications with their friends also seems to have plummeted in recent years. Collectively these trends point toward a need for more media use planning. Such metacognition can allow youth to cultivate their capacity for moderate, constructive digital media practices. That makes promoting and teaching media metacognition is a low cost, and potentially high value, intervention.
Young people experience their use of digital resources in both positive and negative ways. The balance between positive and negative dimensions of digital media use is likely to change as the mix of online behaviors and experiences changes. It is also likely to vary depending on the presence or absence of other factors. Depression and suicidality, for example, amplify negative responses to social media use.
Social Media, Depression, and Suicide
Depression is becoming more prevalent among teens. So is suicide, which is now the second-leading cause of death for that age group. These trends parallel those regarding teen social media use. Except with respect to Facetime, those who use those applications the most are the saddest about how much time they spend doing so. Research has also identified an association (especially for girls) between daily hours of device use and the presence of at least one suicide risk factor. The relationship is not linear, however. When it comes to digital media activities, moderate use may be healthy, but both extremes – isolation and immersion – may be associated with mental health hazards.
Several factors mediate the relationship between social media and depression. Among girls (who spend more time online than boys), unpopularity, lack of self-purpose, and emotional investment are associated with such effects. Other mediating factors include: (i) heavy or excessive use; (ii) devotion to impression management (“social comparison”); (iii) relatively passive use (“lurking”); (iv) fear of missing out; (v) cyberbullying (as perpetrator or victim); and (vi) sexting. Sexting, which is predicted by unsupervised and unrestricted Internet access, is associated with depression and, in most studies, with a substantially greater risk of a suicide attempt. Having more online friends and participating more actively (as opposed to lurking) online both have protective effects. Minority youths, those struggling with their sexual identity, and those with learning disabilities meet others like themselves online. This allows them to feel less lonely and more confident. This is one way that social media use can be associated with increased social support and self-esteem.
The relationship between Facebook use and wellbeing is unclear. There is evidence, however, that Facebook use predicts a mood decline, while interrupting such use predicts a mood improvement. (Face-to-face interactions’ have the opposite effect.) More voyeuristic browsing of others’ curated, idealized versions of themselves – elicits more feelings of envy and mood decline, but no link has been found between more active Facebook use and positive emotional effects.
Especially in this context, the extent to which social media use displaces healthy behaviors (such as face-to-face social contact, school-related activities, physical exercise, and sleep) is a matter of concern. Sleep deprivation strongly predicts depression and suicidality. The bidirectional relationship between depression and addiction makes these data particularly worrisome. Each is a risk factor for the other.
Digital Media for Health Promotion
The vast majority of US teens and young adults are looking online for information about a long list of physical and mental health topics. Health-related mobile phone apps (about things like fitness, nutrition, sleep, menstruation, meditation, mindfulness and mood management) are enormously popular, too. Youth think that they’re capable of understanding, evaluating, and assessing all the health information available to them. They’re not.
Overt and covert corporate messaging (including by seemingly objective celebrity influencers) compounds that challenge. These distorting influences make it imperative that researchers analyze not only how much online health information young people are accessing, but also its character, quality, transparency, and sources.
Fortifying Young Minds Against Misinformation
Even young children are immersed in misinformation now. Children ages 10-18 prefer to get their information from social media, but adults are still kids’ primary sources. This is problematic if those adults are not discerning news consumers. Misinformation almost always elicits a strong emotional reaction, which suppresses objectivity and inspires impulsive re-distribution. This may be why false stories spread much, much faster than true ones. It also may partly explain why, even though people think that they can identify fake news, they can’t. Children particularly lack the skills needed to be critical news and information consumers. Context and confirmability should be informing children’s news consumption and news sharing behaviors. Instead, they are responding to the massive volume of data around them by engaging in headline-deep “information grazing”. This superficial vetting is creating a “bubble effect”. People are relying more and more on fewer and fewer sources, with less and less skepticism.
Emerging Technologies: Children’s Engagement with Intelligent and Interactive Media
Social robot technology is being adopted rapidly. With more than half of US households having Alexa or the like in residence by the end of 2018, we need to think critically about the implications. Kids from preschool age to adolescence know that these devices are artificial, but still attribute to them emotions, thoughts, the capacity for friendship, and a right to moral treatment. We don’t yet know if they will outgrow this view, or if social robots will displace adults and peers as children’s confidants and social and moral mentors. Neither do we know if the independence that children are gaining (by relying on social robots for practical and moral guidance) is a positive or negative cultural change. The answers to these questions are likely to be highly dependent on context. For example, although even high-fidelity mimicry of human-to-human interactions are inauthentic, personified technologies may be useful for interventions with particular children whose particular cognitive sensitivities make actual interpersonal contact too intense.
Digital Aggression and Media Violence
Cyberbullying: What Works, What Doesn’t, and Why Not
A 2011 study showed that both cognitive and emotional empathy measurements declined generationally from 1979 to 2009. This trend is problematic because empathy has psychological and physical benefits. Empathic people tend to be happier, have more meaning in life, feel less depressed, and enjoy certain other health and wellbeing advantages. More specifically, empathy is important because of its association with bullying.
Cognitive empathy (perspective taking) is when people imagine others’ points of view and use body language to determine what’s happening to others. Emotional empathy is empathic concern, or the feeling of compassion and care for others. Bullies have lower emotional (but not necessarily cognitive) empathy than other people do. They may even use their cognitive empathy to target victims. Passive bystanders also show lower emotional empathy. Defenders – people who intervene when they see someone being bullied – score highly on both empathy scales. This difference suggests that empathy training could turn bystanders into defenders. Bullying victims have normal emotional empathy, but slightly below-normal cognitive empathy. They, too, might benefit from training (in how to more effectively take others’ perspective, navigate social situations, and get others to help).
One problem with applying our knowledge of offline bullying to the online type is that the definition of the former, as adapted for the latter, fits poorly. Cyberbullying traditionally is defined as “willful and repeated harm inflicted through the use of computers, cell phones, and other electronic devices.'' This fails to account for power differentials, victim impact, and cumulative effects, and social exclusion. Neither does it adequately distinguish between successful and failed attempts at humor, or accommodate the nuances that surround sexting. Impact, context, and comorbidities matter. Looking forward, increasing cell phone ownership in elementary school children and other trends suggest that cyberbullying researchers should shift some of their attention from adolescents to younger children.
It is as hard for media companies to define bullying as it is for researchers. They also lack information on the context in which the content in question was created, distributed, and received. Algorithmic learning and artificial intelligence technologies are advancing the state of self-regulation, but the process lacks transparency. There is limited evidence regarding its effectiveness, and even that evidence comes from the companies themselves.
Researchers are now testing digital technologies that may positively influence student behaviors in this area. These technologies include an advocative self-reporting app, a prosocial “bystander bot”, a texting-based socioemotional learning program for seventh- and eighth-graders, an avatar-based game; and a partnership with Google and screenwriters on VR scenarios built around an anti-bullying curriculum. Evidence suggests that both bullying and cyberbullying interventions become less effective as children enter 8th
Modern Methods in the Study of Media Violence
The currently available tools for actually visualizing youth media effects have important limitations. The more complicated the behavior or experience being measured, the more difficult it is to interpret an fMRI of the brain engaged in it. Also, fMRI is a better way of imaging the brain’s reaction to general or cumulative effects than momentary ones (such as the different effects of violent and non-violent incidents within a video game). Even when an fMRI reveals a change in the brain’s energy use, it does not explain whether more brain activity is better or worse in a given situation. Doing that takes correlative data from research imaging and clinical and behavioral data from beyond the lab. Electroencephalography overcomes some of MRI’s limitations by offering more flexibility and better temporal resolution. It’s also better at distinguishing between early, primary responses to stimuli and secondary, top-down effects, and it measures brain activity directly (rather than through measure of blood flow and oxygen metabolism). EEG’s disadvantages relative to fMRI include poorer spatial resolution, and its susceptibility to environmental “noise” when used outside of a controlled laboratory environment. Looking at the brain as a dynamic system is an alternative to looking at the activation of specific brain regions. The seed based method and Independent Component Analysis each offer more naturalistic views.
Structural (as opposed to functional) imaging also may shed light on the effect of violent video games. Brain size peaks around age 10-11. Gray matter thins out after that. As for white matter, measurement techniques such as Diffusion Tensor Imaging and Voxel Based Morphometry show a negative association between TV violence and white matter. Connectome imaging, which looks at the connections among brain regions to form a network of hubs, modules, and subnetworks, has not yet been used much to study media exposure.
Imaging, by whatever methods, cannot tell the whole story. It must be combined with other methods, measures, and models, correlated with imaging patients’ clinical data, and interpreted in the context of what we know about brain development across childhood and adolescence. In this research area as in others, we need clear, testable models.
Some hold that the core of the methodological debate over violent media effects is theory. Even the strongest critic of the literature in this field acknowledges that such content has an effect on aggressive cognition and feelings. The idea that something that affects thoughts and feelings doesn’t affect behavior contradicts a basic tenet of psychology. The literature bears this out. Robust findings, across methodologies, have established that playing violent video games is associated with increased aggressive behavior, thoughts, and feelings, and decreased prosocial behavior, empathy, and sensitivity to aggression. Even so, there still are major gaps in the literature regarding gender, ethnicity, and socioeconomic status; dose response relation effects; user motivations; and video game properties (such as first-person versus third-person player perspectives, and cooperation versus competition).
Research shows relatively small effect sizes for violent video games, but looking at the data in different ways can produce much higher correlations. Take more risk factors into account, and extrapolate to the small number of extreme effects, and a child with high exposure to violent games could become 20% or 30% more likely to have a fight during the school year. The data doesn’t show that some populations are more or less vulnerable than others to the effects of violent video games; however, when kids have a lot of risk factors, their risks go up when you add media violence to their lives.
Despite the growing body of evidence regarding the effects of violent media content, controversy persists. This may be because there are methodological, statistical, conceptual, and interpretational reasons why studies underestimate effects or produce non-significant results. There are six principal mistakes being made in the design of randomized controlled studies; six mistakes that undermine cross-sectional and longitudinal media violence effects studies; five statistical errors that further impair media violence effects research; and six main ways in which conceptual and interpretational errors compromise the findings of researchers working in this field. Those who seek to produce valid results need to understand these error types and avoid them. Unfortunately, there are some in this field who are intentionally incorporating these design flaws in their work, in service to a “denialist” agenda
In light of what we already know about the effects of violent video games, the Entertainment Software Ratings Board (“ESRB”) system for rating games requires refinement. The vast majority of games rated for everyone or everyone over age 10 contain characters killing each other. That’s inconsistent with the standard in media effects research, in which violence generally has been understood for the last six decades to mean “characters harming each other”. In light of that definition, violence doesn’t necessarily involved blood, gore, or screaming. (Such things almost certainly increase the physiological desensitization to more of the same, but we don’t know if it increases the effect that killing or harming (within a game) has on aggressive behavior. Teasing apart the variables, and looking at long- versus short-term effects, would take an impossibly large and complicated study.)
The Terminator and Spectator: Is Exposure to Media Violence Linked to Aggression and Violence in the Real World?
Experimental, cross-sectional, and longitudinal studies all show a link between violent video games and aggressive and violent behaviors. Shared definitions of “violence” and “aggression” are needed in order to have a meaningful discussion about that link. The statistical correlation between violent media and aggressive behavior is about 0.2. That’s roughly the same as the average correlation for all social psychology studies over the past hundred years. The correlation between violent media exposure and violent behavior is roughly half that. (Violent behavior, being more extreme, is more difficult to predict.) Even so, that 0.1 correlation is about as strong as the connections between second-hand smoke and lung cancer, lead exposure and IQ decline, calcium and bone strength, and asbestos and cancer.
Meta-analysis showing an increasing association between violent media and aggressive behavior may reflect the increasingly graphic, realistic violence depicted in video games. The more realistic the violence, the larger the effect. However, before age 7 or 8, children struggle to understand both differences between real and make-believe and the motivations informing aggressive behaviors. This rebuts some parents’ belief that their children are unlikely to be influenced by cartoon violence. (Young children’s exposure to the sexualization of female game characters also may be of particular concern.)
ESRB ratings are a double-edged sword. Ratings that indicate highly violent content inform parents, but they also attract children. In any event, ratings cannot substitute for parental participation in children’s media consumption. Such joint engagement creates teachable moments upon which parents must capitalize, because children will perceive parental silence regarding violent content as a tacit endorsement. Parents also need to impose wise constraints on their children’s media consumption habits. As with food, the diet cannot consist of junk. A single exposure to violent content may frighten or upset a child, but repeated exposure reduces these effects and leaves that child craving more.
Problematic Internet Use/Gaming Addiction
Imaging the Loss of Free Will in the Addicted Brain: Implications for Gaming and Internet Addictions
Dopamine is a chemical that modulates all regions of the brain. The basic structural and functional attributes of the brain include particular circuits relevant to addiction, and to the function of dopamine. An unexpected reward triggers dopamine cells to fire. Then, they start to fire when something in the environment suggests that a reward is coming. If that reward is better than expected, then excess dopamine gets dumped into the brain, producing happiness or euphoria and a desire for more. If the brain’s expectation of a reward is frustrated – if it is cued to expect something that doesn’t happen – then dopamine production declines and dysphoria follows.
Drug addiction is associated with a dopamine release deficit. Less dopamine release means more negative symptoms, including inattention, poorer working memory capacity, and poorer probabilistic learning. The adverse neurological effects of addiction are worse in young, developing brains. There is also evidence that certain addictive drugs change how the brain uses glutamate, the synaptic circuitry itself, and the connections among brain regions that operate through that circuitry. (For example, although alcohol use decreases glutamate, chronic use increases glutamate in some regions.)
Because behavioral cycles are similar across all addictions, it is likely that addictions to substances and addictions to behaviors involve similar brain systems. That remains to be confirmed, however. (For example, although it is useful to compare Internet addiction to gambling addiction (also a matter of behavioral, rather than substance, abuse), the literature does not yet include any convincing neuroimaging data with which to compare the two from a neurological perspective. We also need additional data on what addicts’ brains are like before and after addiction to address the question of differential susceptibility.
Who Has a Problem with Internet Use? Screening for, Assessing and Diagnosing Internet Gaming Addiction
There is a conflict in the brains of the addicted, between desires and urges from the limbic system and cognitive control (over these urges and desires) from the dorsolateral prefrontal cortex. Within the addiction process, limbic system impulses become stronger and controls become weaker over time. Psychological and neurobiological considerations suggest that certain factors predispose people to develop and maintain symptoms of Internet use disorders. There is evidence of small to moderate correlations with genetic factors; psychopathological correlates (depression, social anxiety, and ADHD symptoms); and specific personality traits (including high impulsivity). These variables might explain why some people are more vulnerable than others, but not necessarily how any particular individual’s addiction process works. Also, it is unclear whether these identifiable associations reflect causation, in either direction; an interaction between two domains; or just coincident phenomena, neither of which causes the other.
The forms of digital media associated with addictive behavior share certain key characteristics. They are widely accessible, have affordance character, deliver reward, employ intermittent reinforcement, and offer both an escape from the real life’s hazards and opportunities to discover new worlds. Like gaming (and gambling), pornography, shopping and social media-style communication are susceptible to addictive use online. In each case, the term “addiction” should be used cautiously. It must be clinically relevant, offer the best explanation, and be supported by adequate empirical evidence (ideally, from a mix of subjective, interviewing, neuroimaging, and physiological data).
Behavioral addictions (such as Internet addiction) are controversial diagnoses. For example, some scholars find insufficient evidence that gaming disorders should be a formal diagnostic entity while others (including from the World Health Organization) believe the opposite. Due to divergent definitions, and differences with respect to sample sizes, study designs, and analytical methods, estimates for the prevalence of Internet gaming disorder range from 0.3% to 50%. Substantial methodological differences exist even among studies that apply the DSM-5 criteria. One methodological flaw in most of this research is that the studies only count criteria – symptoms – without considering their clinical consequences. Internet gaming disorder’s prevalence declines sharply when the presence or absence of impairment is factored in.
These issues tell us that we need new definitions and more precise, more standardized diagnostic standards. In addition, whether Internet gaming addiction or Internet addiction generally is the proper focus remains an open question with important implications. The best evidence we have (for defining a disorder and estimating functional impairment) is for gaming; however, while gamers mostly are male, those addicted to social network sites mostly are female. This lends a gender equity dimension to the definitional debate. Also, while the Internet can be compared to a drug delivery system, the different activities conducted online are analogous to different substances that people abuse.
Unfortunately, the current structure of the National Institutes of Health, and the allocation of responsibility for addiction studies within it, allow the study of behavioral addictions to fall through the cracks. This is of particular concern because the conflict between academics and public health advocates, on the one hand, and the $100 billion-a-year gaming industry on the other, is a mismatch that parallels the fight over tobacco addiction in the 1990s. The pornography industry also has a $100 billion stake in this fight.
Clinical Solutions for the Youth of the Digital Age
Different clinical responses to digital media use disorders highlight some of the persistent differences among experts in this field. For example, in Germany, a three-phase, standardized cognitive behavioral program (psychoeducation, intervention, and stabilization) has been used to treat 450 patients’ Internet addictions. Their problem behaviors included gaming, social networking, sex, and other online activities. The goals for these patients’ treatment depended on their age. Because the juvenile prefrontal cortex is incompletely developed, adolescent patients can’t self-regulate or apply logic to their behavior, but they demonstrate greater capacity for post-treatment moderation. For these reasons, these younger patients received intermediate training geared toward moderate Internet use. Older patients received psychotherapy combined with abstinence from their problem behaviors.
Another treatment model, used for both adults and adolescents, involves an initial period of nearly complete abstinence from screen use (thought to reset the developing brain to absorb psychoeducation), followed by a slow reintroduction of moderated computer and Internet use. Before proceeding from abstinence to moderate use, patients devise Life Balance Plans that account for how they will reconnect with digital technology and transfer their addictive tendencies. From mid-2015 through mid-2018, two-thirds of 106 adult patients in this program progressed beyond the abstinence phase. Almost three-quarters of those progressed from the second phase to the third, which involved more independent living. Almost three-quarters of those Phase III patients completed the program. Data on why people left the program are incomplete.
Community support and parental involvement are important for patients to succeed in this type of treatment. They tend to have poor social skills, and need remedial social opportunities with peers and professionals who can provide them with a sense of community. Self-monitoring and pro-recovery apps may be useful for highly motivated adult patients already practicing strong recovery behaviors.
South Korean clinical experience casts additional light on this topic. Internet use there has migrated almost entirely onto smartphones. Internet addiction terminology has been replaced with an overdependence paradigm. Abstinence is not the goal now. The treatment objective is to move patients from over-dependence to a more functional, competency- and choice-based relationship with technology.
Digital Media and Developing Bodies
Digital natives are experiencing a wide range of physical difficulties, some of which are associated with serious health issues. For example, obesity is one of the most well-documented outcomes of excessive screen time. It is a global epidemic hitting adolescents, minority groups, and the poor especially hard. It has not been proven that reducing screen time reduces obesity, but at least 15-20 years ago (before the proliferation of smartphones and tablets), data showed that reducing screen time reduced weight gain. The suspected mechanisms at work include changes in eating and sleeping patterns, and possibly changes in physical activity.
Sleep is another area of broad concern. A third of parents of children school-aged or younger report that their kids don’t get enough sleep. Almost a quarter of teens average 6-1/2 hours or fewer of sleep per school night. These statistics have been stable for two decades; however, recent research suggests that since 2010, the percentage of teens getting fewer than 7 hours of sleep a night has increased substantially. The pattern of short sleep during the week and long sleep on weekends is a strong risk factor for such chronic illnesses as depression, anxiety, obesity, and diabetes.
Myopia, which generally develops during the school years and stabilizes in the mid-teens, occurs in about 30% of the US population, but is epidemic in Asian countries. There are predictions that half of the world’s population will be nearsighted by 2050. There is no evidence that near viewing in childhood is associated with becoming myopic, or that intermittent distance viewing relieves eye strain. There is, however, evidence that spending time outdoors does help prevent nearsightedness. This implicates the displacement of healthier outdoor activities by digital, indoor ones.
Musculoskeletal health may be an issue among young technology users, too. When asked, more than half of youth report neck and shoulder pain, and 40% - 50% report hand, wrist, and thumb pain. Repetitive stress and positional overuse are potential risks to be considered. So, too, are distracted walking injuries suffered by people using mobile digital devices.
The Environmental influences on Child Health Outcomes (“ECHO”) Research Program is a seven-year study that began in 2016. ECHO is funded by the National Institutes of Health, and conducted through a large research consortium. It consists of 84 pre-existing pediatric cohorts, representing about 50,000 children and 40,000 mothers and other caregivers in 40 states and Puerto Rico. The oldest cohort’s members were born in 1979. The youngest haven’t been born yet. ECHO is considering how environmental factors ranging from biology to society are impacting five main categories of child health outcomes. Neurodevelopment (from temperament and cognition to autism spectrum disorder) is one of these outcome categories. Another is positive child health, which consists of wellbeing, life satisfaction, happiness, and generally making one’s way in the world. With respect to media in particular, ECHO is addressing basic questions of access and use, and some questions of content. For now, it’s relying on parental reports. Some cohorts are looking more deeply (for example, with media diaries). One cohort is starting at age four months and asking parents about their media use through tablets, and getting very specific about content and co-use.
The Adolescent Brain Cognitive Development (“ABCD”) Study is the largest long-term study of brain development and child health in the U.S. It operates through 21 data collection sites across the country, and a coordinating center in San Diego. Children enter the study at ages 9-10.99. Questionnaires provide baseline data, from children and their parents, on screen time and the specific nature of individual and social digital media activities. Apps for monitoring digital media use are being piloted for incorporation into this study. Future ABCD research will include social networking, environmental sensors, biosensors, and GPS monitoring for subjects over age 18. ABCD already has amended its protocols for Years 1 and 2. It will continue to do so.
Common Sense Media released its latest survey-based national study in September 2018. Over 1,100 US children, aged 13 to 17, participated. Parents also participated, providing information about their own (but not their kids’) media use. (Parents reported spending roughly the same amount of time using media as their children do.) Since the 2012 version of this survey, the percentage of teens using social media has not changed dramatically, but the percentage doing so multiple times per day has increased from 34% to 70%. That coincides with an increase in teens’ reported smartphone ownership (from 41% in 2012 to 89% in 2018). The latest survey also indicates that vulnerable groups are substantially more likely to report negative social media experiences than their less vulnerable peers. In broad terms, they have both heightened positive and heightened negative reactions to social media.
The Common Sense Survey has a number of inherent methodological limitations. Because it is a cross-sectional study, it cannot show causation. Its reliance on teens’ self-reporting makes it subject to several sources of data distortion. Non-response bias is another issue affecting the quality of this data. Also, although the existence of a prior survey (from 2012) creates the appearance of trackable data, these studies did not involve the same participants. As a consequence, the kind of tracking possible in longitudinal research is not possible here. For these and other reasons, teens’ own interpretations of their feelings about social media should not be our only metric. Only valuable resources on these topics include the Pew Research Center’s recent teen social media and technology survey, the Hope Lab study on digital health practices, and the AP NORC survey.
The Cognition Crisis: The Perils and Promise of Technology and the Brain
Human cognition (including attention, memory, perception, emotional regulation, decision-making and reason, imagination, creativity, empathy, compassion, and wisdom) needs to be enhanced. Half a billion people worldwide are debilitated by anxiety, depression, attention deficits, and memory disorders. There is evidence that these symptoms, which have many causes, are on the rise (especially among children). Five problems challenge both the educational and medical establishments as potential means of cognitive advancement: assessment, imprecision, depersonalization, unimodality, and open loop systems.
Human brains are characterized by strong high-level goal-setting and fundamentally limited cognitive control. Distraction is the state in which these two attributes interfere with each other. Technology aggravates distraction; however, it also may allow us to correct this and other deficits. This is because technology increasingly lends itself to the delivery of targeted experiences, which may drive each individual’s brain plasticity in accordance with his or her individual cognitive and behavioral needs.
Taking that step will require more data from closed-loop systems. With each closed loop – each rapid cycle of action, assessment, and adjustment – video game software can acquire a deeper and more comprehensive understanding of the player, and greater ability to target the player’s neural network with constructive stimuli. Such near-real-time data is as necessary for generating therapeutic performance metrics as it is for generating merely engaging ones. Such use of game technology to provide cognitive therapy is being tested now to validate its safety and efficacy.
The Relationship Between Cognition and Media Behavior
We can’t talk about media cognition unless we talk about the different kinds of media interactions. Neither can we talk about cognition as a singular construct. We need to consider how its particular subdomains may be impacted by our variable interaction with media technology. Delayed gratification is one such subdomain of cognition. There is correlational evidence that our digital media habits are producing a society (and more particularly, a society of children) incapable of delayed gratification. We don’t know if media behaviors are changing our inclination to hold out for larger, later rewards, or whether our pre-existing preference for near-term rewards is leading us to more media. Memory is another cognition subdomain. There is compelling, and even causally compelling, evidence that when we are engaged with digital media technologies, we are less able to form memories of our experiences.
Research on habitual “action video game” players found about a one-half standard deviation improvement in cognition over those who don’t generally play a lot of video games. Even in training studies, when you compare subjects forced to play such games against others playing an active control game, there’s a positive cognitive effect better than that produced by most interventions. Attention control is an aspect of cognition that video games impact the most. Some claim that digital media use may adversely impact attention, at least in specific cases, but evidence suggests that this may or may not be true. On the one hand, work on training and multi-tasking indicates that subjecting someone to multi-task training can generalize to broader attentional performance. On the other hand, there also is correlational evidence that heavy media multitaskers perform less well on attentional tasks. A 2009 Stanford study found high media multi-taskers performed worse on various cognitive tasks then their low media multi-tasker counterparts. This proved nothing about causes and effects, but it did show that either (i) media multitasking might cause cognitive problems in some contexts, or (ii) people who already had particular cognitive problems might be more likely to engage in such media consumption behavior. More recent studies found either small effects, uneven effects, for different cognitive tasks, or effects opposite those found in the Stanford study. None of them established any causal connections. A longitudinal study of 1500 Dutch tweens and teens (ages 12-15) attempted to identify such cause-and-effect relationships. A cross-sectional component of the study found that adolescents who multi-task had more attention problems, experienced more distractions in academic contexts, and performed worse academically. They also had more self-reported sleep issues. Longitudinally, a very small effect appeared with respect to distractibility, but only among early adolescents (ages 12-13) and not with respect to grades. The only adverse longitudinal sleep effect was among early adolescent girls.
With respect to attention deficit hyperactivity disorder (“ADHD”) and digital media, it is important to bear in mind that ADHD consists of both naturally occurring individual differences and hyperactivity and impulsivity. It is the latent construct reflected by the nine symptoms of inattention and nine symptoms of hyperactivity with which it is conventionally associated. Risk factors in this area must reflect both the dimensional and categorical perspectives. A recent study of 2500 Los Angeles area high schoolers illustrates this concept. No one in the study initially satisfied the diagnostic criteria for ADHD. Study participants described their participation in 13 different forms of digital media behavior. That information generated scores reflecting the frequency (but not character or content) of each of those behaviors. Reliance on self-reporting, absence of parent interviews, and lack of norms ratings scale limited its predictive power. This data could not predict whether a student without ADHD would develop it. It could, however, predict whether an adolescent would self-report six or more symptoms of inattention and/or six or more symptoms of hyperactivity and impulsivity. High-frequency digital usage positively predicted ADHD symptom criteria status and the number and severity of ADHD symptoms.
To go farther with prescriptions, this field needs to move away from its reliance on self-reporting measures and research based on pro-technology assumptions. That’s challenging, because of the ethical considerations involved in testing anything suspected of having negative effects on children. Help may come from digital media monitoring tools, such as Nick Allen’s Ears and Apple’s newest version of iOS Screen Time. They will provide richly specific objective data about different types of smartphone use.
How to Stop Technology from Destabilizing the World: The Digital Assault on the Human Mind
Content providers’ ability to gather, analyze, and deploy information about our tastes and tendencies gives them an insurmountable advantage over sovereign human choice. This allows them to acquire and exploit influence. The details of the algorithms with which they do so is beyond the comprehension of even their creators. It seems, though, that these attention-grabbing algorithms are introducing an intrinsic bias toward sensationalization into the world – “a race to the bottom of the brain stem”. This tilt of the playing field toward radicalizing content introduces into people’s lives a thread of paranoia that extends into the social fabric. It may not be an intentional consequence of social networks’ design, but it is a predictable (and exploitable) one. Children are particularly vulnerable. All of us have cognitive limits, but children’s self-worth, identity development, and capacity for discernment are limited, too.
We don’t need more research to decide to protect children. That doesn’t mean insulating them from technology. It means guiding them toward technology that makes developmental sense for them, rather than the tech that’s just intended to capture maximum attention. Itemize the public harms that result when companies privately profit from people’s attention would be a step toward positive change. Increased public awareness and class action lawsuits also might help. Public concerns already seem to be influencing corporate behavior at Facebook, Apple, and Google. Alternative business models, which better reflect the value of people as the product in social media transactions, may induce constructive cultural change.
Cell Phones in Schools
Survey data indicate that 89% of people ages 12-17 now have smartphones. There are many ways to the role of cell phones in schools. For example, in September 2018, the French government banned mobile phones from schools for students from preschool through age 15. New York City banned phones from schools for ten years, but relented in 2015 in the face of pressure from parents who wanted to be able to reach their kids. Anecdotes from midwestern school offer other alternatives. Some impose uniform, campus-wide policies. Others give teachers more autonomy in their own classrooms, give more latitude to older students, or distinguish between uses in and out of class. Still others distinguish among different kinds of use (such as non-consensual recording or cyberbullying). Educators in schools with one-computer-per-student programs don’t see a role for phones in their classrooms. Where not every student has a computer, permissive phone policies may give necessary access to educational and administrative information.
A British study in 2015 compared secondary students’ standardized test scores before and after the adoption of cell phone bans. They found that if a school banned cell phones and got its students to comply, then student test scores improved. The biggest driver of this effect was improvement by the schools’ most underachieving students. There was no effect on students who were top performers before they had to give up their phones at school.
A current study of 2100 North Carolina teens and tweens (ages 10-15) poses a number of questions about the effects of owning and using phones. About half of the participants under age 12 already had their own phones. That percentage was around 75% by age 13, and around 85% for those 14 and older. For each age group, approximately the same percentages had their own social media accounts. After controlling for gender, race, family and neighborhood income, age, and other covariates, there were no associations between owning a mobile phone and either test scores or conduct problems. On the other hand, having a social media account was associated with lower standardized math test scores, more conduct problems, more psychological distress, and lower general health. The effect sizes were very small, but across a large population. Economically disadvantaged kids were twice as likely to report perceived impairments from phone use as more affluent ones were. This suggests a particular kind of differential susceptibility; namely, that social media is exacerbating vulnerabilities arising from racial, economic, and other differences among children.
A novel study of phones in school capitalized on the idea that ostensibly instructional technologies can be evaluated using a course-embedded experimental paradigm. A course was organized so that students’ substantive academic performance would reveal the educational impact of allowing, or prohibiting, electronic devices in a college class. The results can be explained in light of what we know about divided attention and cognition. Divided attention affects cognition in several ways. One immediate effect is that when you respond to one thing, you miss something else. Another effect is less immediate. Learning is a byproduct of action. People remember the targets of their action, the actions themselves, and the consequences. Study and learning constitute mechanism of targeting, action, outcome, and memory -- “rehearsal”. During a divided attention task, rehearsal is suppressed. This has an insidious long-term effect on retention. The farther out you go in time from the divided attention task, the greater the adverse memory effect.
This reasoning not only explains why students with free access to their electronic devices during class underperformed there and on subsequent exams. It also explains why students who used their devices to look up homework answers (rather than work out those answers themselves) excelled on that homework but generally underperformed on their exams. A growing percentage of students are Googling their homework, and getting no educational benefit from it. If the decade-long trend in this single experiment is part of a broader phenomenon, then finding homework answers online will soon be so prevalent that homework will have no value at all. Traditional instructional materials will be useless. Some students will still excel on exams by cramming, but they will simply suffer more, achieve briefly, and retain less than similarly capable students of past years.
Technology and Laptops in Schools
In contrast to cell phones, laptops are the technology that we invite into classrooms because (to paraphrase Seymour Papert) they are the prime instruments for intellectual work in our times. That having been said, the opportunities and challenges of digital learning involves more than just laptops. They encompass various devices and online resources, at school and elsewhere, and opportunities for customizing content, individualizing instruction, and transforming the role of teachers.
Recent data from Dallas and Milwaukee sheds light on this complexity. There is evidence of some success (for example, reading achievement in Dallas elementary schools using tablets, especially in bilingual classrooms). There are everyday problems, too, with potential long-term adverse effects. For example, technology failures are consuming valuable instruction time. The more time lost to tech issues, the lower students’ predicted scores on standardized reading and math exams.
The use of technology to provide alternative, online classes for kids failing in conventional classroom settings raises particular concerns. This practice is moving kids – many of whom have underlying reading deficiencies or learning disabilities – into rooms with high student-teacher ratios, little substantive help from teachers, and standardized online instructional material designed for passive learning. Such an environment lends itself to time off task, ranging from simple inactivity to phone use. This may help to explain why the more years of online courses high school students take, the worse their grade point averages and math and reading test scores, and the fewer their credits earned.
Neuroscience sheds light on this problem. The brain regions responsible for higher cognitive functioning continue to develop through adolescence. “Conversational turns” (dyadic signaling) between adults and developing children are important for this reason. Such interactions strengthen connections between developing brain regions critical for language. This process may be especially important to children whose classroom struggles are associated with below-grade reading ability or learning disabilities.
Evidence from the Maine Learning Technology Initiative sheds further light on this subject. In the early 2000s, Maine launched the MLTI to give every seventh- and eighth-grader in the state, their teachers and their administrators a laptop computer. It also wired their schools and provided teacher and leader professional development. This program expanded through Grade 12 in 2009. Broadly speaking, the purpose of the MLTI is not just to give kids computers, but to accelerate their learning. There is little evidence that it’s succeeded. Students’ and schools’ responses to their new technology have largely reflected their socioeconomic status.
Together, these studies and stories remind us that it is not enough to buy the tools. Implementation is key. Investments in classroom technology must be accompanied by proportional investments in initial and ongoing professional development and on-site tech support. Digital tools and programs must be tailored to the individual needs of the students being directed to use them. This is especially true for students with special educational needs. Inappropriate technology implementation actually may be exacerbating, rather than resolving, inequities.
In addition to bringing hardware and software resources into schools, it may be valuable to import some of the youth technology culture. School culture has long been directed toward optimizing reading and math. This differs significantly from the culture of the digital devices that young people use outside of school. The increasing penetration of those devices into the school space invites a question. Looking at how kids learn with technology outside of the classroom, are there things other than reading and math that tech can help them master inside the classroom? This is not a traditional question of media consumption or media exposure. It is a reconsideration of what learning is, in terms of four ways in which young people interact with media. According Mimi Ito and her colleagues, these are: relationships, tinkering, making, and circulation. Virtual spaces are supplanting physical environments for the safe conduct of relationships. Tinkering involves manipulating the technologies through which those relationships now happen. It often leads to making – creative activities ranging from coding to content production to 3-D printing. Circulation is engagement in a core activity with concurrent social interaction. These non-school activities illustrate 21st century models for learning, which may have a place in 21st century schools.
Compare, for example, school-based coding education (through courses and degree programs available in only 10% of schools) to how people learn in a popular mobile storytelling app. To join the app’s community, you must learn to code, and then do it. There are no grades or mandates. The inducement is that learning to code secures opportunities to tinker, to make, and to circulate personal creations, and become of a community. Other environments in which young people are using digital media to develop real life skills include Chicago’s Youmedia program, within social games and in in-person “maker spaces”. Such examples suggest that making the line between life technologies and learning technologies a little more porous may help schools to use digital media for learning (rather than just for replicating their traditional outcomes).
Won’t Somebody Think of the Children?! Examining Privacy Behaviors of Mobile Apps at Scale
Every device, app, and website through which children access digital content has attributes that, collectively, reveal things about its users and the substance and circumstances of their online activities. The ways in which games, devices, and networks talk to each other generates, exploits, and reveals some of that information. Content makers, advertisers, and data brokers occupy a personal information ecology through which information is collected, cross-referenced, and aggregated into actionable personal information (including about children). Industry policies and regulatory schemes exist that seem intended to protect children from the collection, use, and misuse of their personal information are effectively serving those purposes. Incomprehensible and intentionally vague corporate privacy policies, software development kits intended for use in adult products and misused for kids, weak and infrequent enforcement action by federal and state officials, and a robust market for monetizable digital user profiles are collectively depriving children of reasonable protection in this regard. Even the Safe Harbor mechanism under the Children’s Online Privacy Protection Act seems not to be working in this respect. Research indicates that Safe Harbor-certified products are as likely as others to function in non-compliant ways.
Only the Federal Trade Commission (“FTC”) and state Attorneys General are authorized to bring civil enforcement actions under COPPA. They cannot bring criminal actions for COPPA violations, and at least the FTC acts only on consumer complaints. Between resource constraints and the difficulty that consumers have even identifying how, when and by whom their privacy has been violated, this scheme has little deterrent effect. A private right of action, and the threat of class action suits, might change things. The relatively new European General Data Protection Regulation may prove more effective, but it remains to be seen how that will be enforced.
Measurable effects in the social sciences tend to be small in absolute terms. Their significance depends in part on the size of the population affected. This is one reason why, to accurately answer the questions they pose, youth media effects studies must look at populations that are large enough, and ethnically, geographically, and socioeconomically diverse enough, to produce broadly applicable and statistically robust findings.
Cross-sectional studies cannot reveal causal connections between phenomena. They are appealingly simple, and useful for identifying associated phenomena, but they lack explanatory power. Studies that rely on surveys and self-reporting also have problematic limitations. They do not measure actual stimuli and responses. They measure only what people remember or admit. Collectively, these types of research identify associations that may be causal or coincidental. To determine which is the case, and to identify specific causes, effects, and the mechanisms that connect them, will require more longitudinal studies in naturalistic settings.
In contrast to such indirect study methods as surveys, questionnaires, journals, and interviews, emerging technologies offer new ways to measure what’s actually happening in children’s media environments. Some of these technologies already are gathering digital media and digital device user information for commercial purposes. This makes accessing industry data, and deploying clinical imaging and monitoring tools, for research purposes a priority. Without such direct assessment data, even robust longitudinal studies will not be enough to precisely establish actionable, causal relationships between digital media and devices and child development outcomes.
Even as we seek new tools and methodologies, we must make more effective use of existing ones. One way to fill gaps in our knowledge is through secondary analysis of archival materials and findings from adjacent fields of study. Data gathered by other research communities, for other purposes, may tell us important things about the relationships between children’s media-related experiences and behaviors and their cognitive and psychosocial development.
This elegant use of diverse information sources fits well with a “whole child” approach to youth media effects research. Only by looking at the whole child can we gain a usefully comprehensive understanding of how digital media affect developing minds. Answers about attachment and emotional regulation, executive function, social development, language acquisition, other determinants of child and adolescent mental health, sleep, physical health and wellbeing, weight management, play, and all the different forms of media to which children are exposed, each represent only pieces of the larger puzzle that we seek to solve.
We also must bear in mind that the purpose of this work is not to score points, win arguments, or secure patronage from industry groups. It is to help children. We must neither presume the answers to our questions, nor allow ideological or financial partisanship to disrupt the thoughtful, respectful, collaborative search for truth.