The Other Side of the Story: Why Evidence Based Child Protection Requires More Than One Side of the Research
CAVEAT: This article is not intended to dismiss, minimize, or question the very real harms that “some” youth and teenagers have experienced through technology, the internet, and social media. Here at The White Hatter, we have worked directly with families whose children have experienced cyberbullying, sextortion, online sexual exploitation, grooming, and other forms of technology-facilitated victimization. We understand firsthand how devastating these experiences can be, not only for the young person involved, but also for their family. We also strongly support holding technology companies accountable for platform designs, algorithms, and business practices that expose children to foreseeable harms. Our concern is not that research identifying these harms is being shared publicly, because it absolutely should be. Rather, our concern is that some public discussions about youth, technology, and social media appear to emphasize research supporting a particular conclusion while giving considerably less attention to credible scientific findings that challenge, qualify, or complicate that conclusion. We also recognize that this concern applies equally to those who emphasize the benefits of technology while minimizing evidence of harm. Scientific integrity requires us to examine research that challenges our own beliefs, not just the beliefs of others. The central question we explore in this article. “Are parents, caregivers, educators, and policymakers getting the full evidence based research picture when making decisions about youth, technology, and social media?”
Much of today’s public conversation about youth, technology, the internet, and social media has become increasingly focused on the negatives. From attention grabbing headlines and emotionally compelling documentaries to social media campaigns advocating delayed access to smartphones and social media, parents, caregivers, and educators are frequently exposed to messaging suggesting that technology is a significant driver of declining youth mental health. We recognize that many of these concerns are motivated by a genuine desire to protect children, a desire we share. However, through our work reviewing scientific research, participating in public discussions, and engaging with parents, caregivers, educators, and fellow advocates, we have observed that studies identifying risks and harms often receive considerable attention, while research reporting weaker associations, methodological limitations, or more complex relationships does not always receive the same consideration. We acknowledge that this observation does not establish how frequently selective reporting occurs across all media, documentaries, advocacy campaigns, or public presentations, something that would require a systematic examination of how research is being communicated. Nevertheless, we believe it raises an important question that deserves greater public attention, “Are parents, caregivers, educators, and policymakers consistently receiving a sufficiently complete and balanced understanding of the scientific evidence before being encouraged to support particular parenting philosophies, educational policies, or legislation?”
When research identifies an association between social media use and adolescent depression, that finding deserves serious consideration. However, so does research identifying little or no association, longitudinal studies suggesting that psychological distress may sometimes precede increased social media use, and research examining the influence of family relationships, childhood adversity, peer connections, and pre-existing vulnerabilities. What about studies identifying meaningful benefits associated with certain online experiences, including social connection, identity exploration, access to information, educational opportunities, and access to supportive communities? These findings do not automatically disprove research identifying harm, nor should they be used to dismiss the experiences of young people who have suffered genuine online victimization. However, they may provide important context for understanding when, how, why, and for whom technology use is associated with particular outcomes. We believe all of these findings deserve consideration, particularly when research is being used to justify broad recommendations or public policies that could affect millions of young people.
There is a commonly used term for selectively emphasizing research that supports a preferred conclusion while overlooking credible evidence that challenges it, and that term is “cherry-picking the evidence.” This practice can occur on either side of the technology debate. Someone arguing that social media is inherently harmful can engage in cherry-picking, just as someone arguing that concerns about online harms are exaggerated can do the same. The problem is that selective evidence can create the appearance of scientific certainty where meaningful questions, limitations, and disagreements remain. When research is presented primarily to support a predetermined conclusion rather than to improve our understanding of a complex issue, parents and caregivers may be left with an incomplete picture of what the science actually tells us. We believe there is an important difference between using research to help people understand an issue and selectively presenting research to persuade people to accept a particular position. The former encourages informed decision making, while the latter risks allowing advocacy to determine which scientific findings receive attention.
Why We Compiled 47 Research Studies and Reports
It was this concern that motivated us to compile and share 47 research studies and reports published between 2018 and 2026 examining relationships between technology, social media, educational screen use, adolescent development, and mental health. Our objective was straightforward, to bring greater public attention to credible research that we believe deserves more consideration in discussions about young people and their use of technology. Some of these studies challenge widely repeated assumptions about screen time and adolescent mental health, while others examine whether smartphone ownership or general social media use independently predicts psychological difficulties. Several highlight the importance of family relationships, childhood adversity, peer connections, developmental differences, and the broader environments in which young people grow and develop. Other research identifies limitations in drawing causal conclusions from studies reporting associations between technology use and mental health outcomes. Some findings suggest that these relationships may operate in both directions, meaning that young people experiencing psychological distress may increase certain forms of technology use, while particular online experiences may also contribute to distress. Importantly, some research in our collection also identifies legitimate risks, problematic patterns of technology use, and associations with poorer psychological outcomes. We included these findings because our intention is not to replace one selective narrative with another.
We want to make an important distinction clear, our collection of 47 studies and reports was deliberately assembled to highlight credible research that we believe receives insufficient attention in some public discussions. It is not a systematic review of the entire scientific literature, nor is it intended to represent a statistically representative sample of all published research on youth, technology, and mental health. We are also not suggesting that these 47 studies outweigh, invalidate, or disprove research identifying meaningful risks and harms. Our collection includes different types of evidence, including longitudinal studies, randomized controlled trials, scientific reviews, meta-analyses, preprints, and institutional reports. These sources do not all carry the same evidentiary weight. Some have undergone formal peer review, while others have not. Some examine large populations over time, while others investigate specific interventions, behaviours, or outcomes. We believe these distinctions are important because the number of studies supporting a particular position is not, by itself, a reliable measure of scientific consensus. Forty seven studies do not automatically outweigh ten studies, just as ten studies do not automatically outweigh forty seven. Research quality, methodology, sample characteristics, effect sizes, limitations, and relevance to the question being examined all matter.
Our collection is intended to supplement the broader scientific literature, not replace it. We encourage readers to examine these studies alongside research documenting significant online harms and to approach both bodies of evidence with the same critical thinking. We also recognize that scientific evidence does not always provide simple or definitive answers, particularly when examining complex relationships involving human behaviour, adolescent development, family environments, and rapidly changing technologies. This is precisely why we believe a broader presentation of the research is so important. Our goal is not to prove that one side of the debate is correct, but to encourage a more complete understanding of what the available scientific evidence does and does not establish.
What Does the Research Actually Tell Us About Youth, Technology, and Mental Health?
One of the most important lessons emerging from research examining youth, technology, social media, and mental health is that the relationship is more complex than the often repeated narrative that social media automatically equals harm. Credible research has identified genuine risks, documented serious harms, and reported associations between certain patterns of technology use and poorer psychological outcomes. Other credible research has identified relatively small average associations, inconsistent findings, limited evidence of causality, and important differences between individual young people. These findings are not necessarily contradictory. They may reflect differences in what researchers are studying, how technology use is measured, which outcomes are examined, and the characteristics of the young people participating. For example, a study measuring the total number of hours teenagers spend using screens is not necessarily examining the same question as a study investigating exposure to cyberbullying, compulsive social media use, or algorithmically recommended harmful content. Similarly, research examining general social media use may produce different findings from research examining specific platforms, activities, or online experiences. This distinction matters because we should not assume that all screen time is equivalent, that all social media experiences are interchangeable, or that every young person responds to technology in the same way.
One of the most important principles in scientific research is that identifying an association between two variables does not automatically establish that one causes the other. For example, if researchers find that teenagers who spend more time on social media also report higher levels of depression, several explanations may be possible. Social media use may contribute to psychological distress, young people experiencing psychological distress may turn to social media more frequently, or other factors, including family difficulties, peer relationships, sleep disruption, previous victimization, or existing mental health challenges, may influence both technology use and psychological well being. More than one explanation may also be operating simultaneously. This does not mean that research identifying associations should be dismissed. Rather, it means that we must be careful about the conclusions we draw from it. Longitudinal studies can help researchers examine how relationships develop over time, while randomized controlled trials may provide stronger causal evidence about specific interventions. However, no single research design answers every question, and even well-designed studies have limitations that should be considered when interpreting their findings.
We also believe it is important to clarify what we mean when describing the research as mixed. Mixed findings do not establish that social media is harmless, just as research identifying an association between social media use and psychological distress does not automatically establish that social media is the primary cause of that distress. A study that does not identify a statistically significant relationship does not necessarily prove that no meaningful effect exists. Likewise, a relatively small average association across a large population does not rule out substantial negative effects for particular young people. Some teenagers may experience serious psychological distress associated with particular online experiences, while others may experience meaningful benefits or little measurable impact. Small average effects may also have public health significance when they affect large populations. This is why responsible research interpretation requires more than simply asking whether a study found a positive, negative, or statistically significant result. We should also examine the size of the effect, the uncertainty surrounding it, the quality of the research, and whether the findings have practical significance.
Another important distinction often lost in public discussions is the difference between research examining general adolescent mental health and research documenting specific forms of online victimization. These are related but different scientific questions. For example, research finding a weak average association between general screen time and depression does not negate the serious psychological consequences that can result from cyberbullying, sextortion, grooming, or online sexual exploitation. As stated earlier, we have worked directly with families who have experienced these harms, and we understand that their consequences can be devastating. However, evidence documenting these specific harms does not automatically establish that a universal social media age restriction, smartphone ban, or delayed access strategy will prevent them. Research identifying cyberbullying as a significant risk provides a compelling reason to address cyberbullying, but it does not automatically demonstrate that banning social media accounts for everyone under 16 will reduce cyberbullying, particularly if harmful behaviour shifts to other digital environments. That possibility requires investigation rather than assumption. Similarly, research finding limited population level mental health effects from general technology use does not justify dismissing the need for stronger safeguards against online exploitation.
One of the limitations of relying exclusively on population level research is that average findings do not necessarily describe the experiences of individual children. Even where researchers identify relatively small average associations between social media use and psychological well being, some young people may experience substantial harm, while others may benefit from online social connections, educational opportunities, or access to supportive communities. A child’s developmental stage, maturity, family environment, previous experiences, peer relationships, and existing vulnerabilities may all influence how they interact with technology. These factors do not mean that particular groups of children will inevitably experience harm. Rather, they remind us that young people are not a homogeneous population. A 13 year old with strong family support, established boundaries, and age appropriate digital literacy skills may have a very different online experience from another 13 year old facing significant psychological distress, social isolation, or previous victimization. Chronological age can be an important consideration, but it does not capture every factor relevant to readiness, risk, or resilience. This is one reason we have consistently encouraged parents and caregivers to consider their individual child’s circumstances rather than relying exclusively on universal rules.
We Cannot Wait for Perfect Evidence Before Protecting Children
Here at The White Hatter, we recognize an important argument advanced by advocates supporting stronger restrictions on youth access to smartphones and social media, including proponents of the Delay Is the Way philosophy. That argument is straightforward, when credible evidence identifies potentially serious risks to children, waiting for definitive scientific proof before taking protective action may expose young people to preventable harm. We believe this is a legitimate position deserving serious consideration, and we agree with its central premise. We cannot allow the absence of perfect scientific evidence to become an excuse for doing nothing when children are experiencing meaningful and foreseeable harm. We have worked with families whose children have experienced devastating online victimization, and we understand that telling those families to wait several more years for better research is neither a practical nor compassionate response. Precautionary approaches have an important role in child protection, particularly when the potential consequences of inaction may be serious.
However, acknowledging the need for precaution does not automatically establish that every proposed intervention is necessary, proportionate, effective, or free from unintended consequences. This is where we believe the public conversation needs to become more thoughtful. The question should not simply be whether we act now or wait for better evidence, the more important question is, “What actions are most likely to reduce harm, what evidence supports those actions, and how we will determine whether they are working.” Reasonable people can agree that children need stronger protection while disagreeing about which protective measures are most likely to achieve that objective. Such disagreements should encourage better research and more rigorous evaluation rather than accusations that those questioning a particular intervention do not care about children’s safety. We can support stronger child protection while also asking difficult questions about the effectiveness of proposed solutions, and those positions are not mutually exclusive.
We also believe the strength of evidence required to justify an intervention should be considered alongside the seriousness of the risk and the potential consequences of the proposed response. Not every protective measure requires the same level of scientific certainty. Introducing stronger privacy protections, improving reporting mechanisms, or restricting clearly inappropriate contact between adults and children may be reasonable precautionary steps even when every potential outcome has not been conclusively studied. However, broader interventions that significantly restrict access, affect large populations, or create substantial privacy and accessibility implications deserve particularly careful examination. This does not mean such interventions should automatically be rejected, it means that the justification for them should be proportionate to their potential consequences. We must also consider the risks of inaction, because failing to address harmful platform features or known patterns of online victimization may expose young people to preventable harm. Introducing restrictions without adequately considering their consequences may also create new challenges, and neither possibility should be ignored.
Evidence That a Problem Exists Is Not the Same as Evidence That a Particular Solution Will Work
This is perhaps the most important distinction we want parents, caregivers, educators, and policymakers to understand. Identifying a problem and demonstrating that a proposed intervention will solve it are two separate scientific questions. Research documenting cyberbullying, sextortion, online sexual exploitation, problematic social media use, harmful platform design, or psychological distress provides compelling reasons to take protective action. However, that same research does not automatically demonstrate that a universal age restriction, smartphone ban, or delayed access strategy will produce the intended improvements in youth safety or mental health. Establishing that a problem exists is an important first step, but determining whether a particular solution is effective requires additional evidence examining that intervention and its outcomes. Too often, we believe, these two questions are treated as though they are interchangeable.
For example, introducing a social media age restriction for everyone under 16 may appear to be a decisive response to concerns about adolescent mental health, and it may reduce exposure to certain risks for some children. However, whether such a restriction improves population level mental health outcomes, reduces specific forms of online victimization, or creates unintended consequences is an empirical question requiring evaluation. The popularity of a policy does not establish its effectiveness, nor does the sincerity of those advocating for it. The same principle applies to school smartphone restrictions. A school may introduce a policy limiting personal smartphone use during instructional time to reduce distraction and improve learning, and that may be a reasonable objective. However, the success of the policy should ultimately be assessed by whether it produces measurable improvements in the outcomes it was intended to address. A reduction in visible smartphone use may demonstrate compliance with the policy, but it does not automatically establish improvements in academic achievement, mental health, or student well being.
We should also recognize that not all restrictions are equivalent. A parent’ or caregiver’s decision to delay a child’s access to a particular social media platform is fundamentally different from a school’s restriction on personal smartphone use during instructional hours. Both are different from government legislation establishing a minimum age for social media accounts. Each intervention operates in a different environment, pursues different objectives, and may produce different outcomes. Evidence supporting one approach should not automatically be interpreted as supporting another. For example, research demonstrating that restricting smartphone use during classroom instruction improves attention would not necessarily establish that prohibiting social media accounts for everyone under 16 improves adolescent mental health. Similarly, evidence supporting parental involvement and gradual access to technology does not automatically establish that every child benefits from earlier access. Each intervention must be evaluated according to its specific objectives, implementation, and outcomes.
Delaying Access and Teaching Digital Literacy Should Not Be Competing Philosophies
One concern we have with some public discussions is the suggestion that delaying access to technology represents the primary solution to the challenges young people face online. We recognize that delaying access may be an appropriate decision for some children and families. Depending on a child’s age, maturity, circumstances, and available supports, parents and caregivers may reasonably decide that certain platforms or technologies are not yet appropriate. We support their ability to make those decisions. However, delaying access and teaching digital literacy should not be treated as mutually exclusive strategies. Parents and caregivers can delay access to particular platforms while simultaneously helping their children develop the knowledge, judgment, and skills needed to navigate digital environments. These skills can be developed through age appropriate experiences, supervised technology use, family conversations, educational programs, and gradual increases in independence. Our concern arises when delay is presented as a complete solution rather than one possible protective measure within a broader developmental approach.
Delaying access does not automatically teach a young person how to recognize manipulation, identify misinformation, establish boundaries, respond to unwanted contact, protect personal information, or seek help when something goes wrong. These are skills that require education, practice, guidance, and supportive relationships. We also need to consider whether restrictions may shift certain behaviours to other online environments, including gaming platforms, messaging services, or emerging technologies such as social AI. Such displacement should not be assumed to occur in every case, but it is a possible outcome that deserves investigation. At the same time, we must be willing to apply the same scientific scrutiny to digital literacy education that we apply to age restrictions. Teaching digital literacy does not automatically guarantee safer behaviour, just as increasing awareness of online risks does not necessarily prevent victimization. Strong family relationships and parental involvement may provide important protective benefits, but they cannot eliminate every risk, particularly when harm results from offender behaviour or platform design. If we expect advocates of age restrictions to demonstrate that their preferred interventions work, we must be prepared to demonstrate the effectiveness of the approaches we support as well.
Another concern we have is that much of the public conversation focuses on when children should gain access to technology rather than what they are gaining access to, and we believe both questions matter. The design of digital environments can influence how young people interact with technology, what content they encounter, how their personal information is collected, and whether they are exposed to unwanted contact or other foreseeable risks. This is why we have consistently advocated for meaningful corporate accountability. Children and families should not bear sole responsibility for navigating digital environments where potentially harmful features or business practices could be addressed through safer design. Age appropriate privacy settings, stronger protections against unwanted adult contact, transparent recommendation systems, accessible reporting mechanisms, and restrictions on manipulative design practices are all areas deserving serious consideration. However, these approaches should also be evaluated rather than automatically presumed effective. The broader objective should be to create safer digital environments while equipping young people with the skills and support necessary to navigate them. Protecting children requires shared responsibility between technology companies, governments, schools, parents, caregivers, and the broader community.
How Will We Know Whether the Interventions We Introduce Are Actually Working?
This is where we believe the discussion about protecting children needs considerably more attention. If governments, schools, technology companies, or advocacy organizations introduce interventions intended to improve youth safety or mental health, how will we determine whether those interventions are successful? Too often, we believe, success risks being measured by whether a policy was implemented rather than whether it achieved its intended outcomes. A government announcing an age restriction may represent the achievement of a legislative objective. A school introducing a smartphone ban may represent the implementation of an educational policy. An advocacy organization persuading families to delay social media access may represent the success of a public awareness campaign. However, none of these outcomes, on their own, demonstrates that children are safer, healthier, more resilient, or better prepared to navigate their onlife world. We believe meaningful evaluation should begin by clearly identifying the problem an intervention is intended to address, establishing measurable outcomes, and determining how those outcomes will be assessed over time.
For example, if the objective of a social media restriction is to reduce cyberbullying, we should examine whether cyberbullying decreases following implementation. If the objective is to improve adolescent mental health, we should evaluate relevant psychological outcomes rather than simply measuring how many young people have been prevented from opening accounts. If the objective of a school smartphone policy is to improve classroom learning, we should examine educational outcomes rather than assuming that reduced device use automatically translates into better academic performance. Importantly, we should distinguish between changes in reported incidents and changes in actual prevalence. An increase in reports of online victimization may sometimes reflect improved awareness and reporting rather than an increase in victimization itself. Similarly, a reduction in reports does not necessarily mean that harmful behaviour has disappeared. This is why choosing appropriate outcome measures and interpreting them carefully is essential.
We should also recognize that changes observed after a policy is introduced cannot automatically be attributed to that policy. Other factors may be contributing to those changes, including broader social trends, changes in family circumstances, economic pressures, educational environments, or the evolution of technology itself. Where feasible, independent researchers should use appropriate comparison groups, baseline measurements, longitudinal follow-up, and other rigorous methods to determine whether an intervention is responsible for the outcomes being observed. We should also examine whether benefits differ across age groups, developmental stages, or other relevant circumstances. An intervention that benefits one group of young people may have limited effects or unintended consequences for another, and those differences deserve consideration rather than being obscured by population averages.
Another important consideration is the possibility of unintended consequences. An intervention may achieve one objective while creating other challenges. For example, restrictions could potentially influence access to supportive online communities, create accessibility barriers, encourage migration to less regulated platforms, or introduce new privacy concerns through age-verification systems. These are possibilities to investigate, not outcomes that should be presumed inevitable. Equally important, failing to introduce effective protections may allow preventable harms to continue. A responsible evaluation should examine both intended benefits and unintended consequences, recognizing that every policy decision involves trade offs, including the decision not to act. The objective should not be to identify a policy that carries absolutely no risk, because such a standard would be unrealistic. Rather, the objective should be to determine whether the anticipated benefits of an intervention justify its limitations, costs, and potential consequences when compared with reasonable alternatives.
Perhaps most importantly, we should be willing to modify our approach when new evidence becomes available. If an intervention is demonstrably effective, we should consider how to strengthen or expand it. If the results are mixed, we should investigate why. If an intervention is ineffective or produces unintended harms that outweigh its benefits, we should be prepared to change course. This should not be viewed as a policy failure, but rather as evidence informed decision making in action. The purpose of evaluation is not to defend the intervention we originally supported, it’s to determine whether that intervention is delivering the outcomes children deserve. We believe this distinction is especially important when discussing policies that may affect an entire generation of young people, because the consequences of ineffective or poorly designed interventions may not become apparent until years after they have been introduced.
We Can Act Now and Continue Learning
We reject the idea that protecting children requires choosing between immediate action and better scientific evidence. These objectives can, and should, work together. Where credible evidence identifies meaningful and foreseeable risks, reasonable protective measures can be introduced while researchers simultaneously evaluate their effectiveness. This allows us to respond to emerging concerns without treating our initial assumptions as permanent conclusions. Some outcomes may be measurable relatively quickly, while others, particularly those involving long term adolescent development or mental health, may require years of follow up. We should also recognize that the digital environment is constantly changing. Platforms evolve, new technologies emerge, and young people adapt their online behaviour. An intervention that appears effective in one environment may not produce the same results under different circumstances. This is why ongoing research and independent evaluation are so important, and why we should not expect every policy to remain unchanged simply because it was once considered the best available response.
We should also recognize that scientific evidence, while essential, is not the only consideration involved in public policy. Governments must also consider children’s rights, privacy, accessibility, parental autonomy, educational responsibilities, and their legal and ethical obligations to protect young people. Scientific research can help establish what is happening, what may be contributing to harm, and whether particular interventions are effective, but it cannot independently determine how competing rights and values should be balanced. These considerations reinforce the importance of transparent, proportionate, and accountable decision making. Evidence informed child protection should be an ongoing process of action, evaluation, learning, and adjustment rather than a one time decision based on the best information available at a particular moment.
We should never allow the absence of perfect evidence to justify doing nothing. However, we should be equally cautious about allowing the urgency to do something to replace the responsibility to determine whether what we are doing actually works. Both action and inaction can have consequences, and both deserve thoughtful consideration. We believe the most responsible approach is one that acknowledges uncertainty, responds proportionately to credible risks, and remains willing to change as stronger evidence becomes available.
Evidence-Based Should Mean Applying the Same Standards to Everyone, Including Ourselves
One of the principles we have consistently emphasized at The White Hatter is that being evidence based requires more than finding research that supports a particular belief, philosophy, campaign, or policy position. It requires being willing to examine credible findings that challenge our assumptions, acknowledge methodological limitations, recognize uncertainty, and change our position when stronger evidence becomes available. This applies to those advocating for universal age restrictions, those opposing such restrictions, technology companies promoting the benefits of their products, educators developing school technology policies, and policymakers introducing legislation. Importantly, it also applies to us here at The White Hatter. We believe digital literacy education, supportive family relationships, age appropriate boundaries, safer platform design, and meaningful corporate accountability are important components of protecting young people. However, we should not assume that every educational intervention, parenting strategy, or platform safeguard will automatically produce the outcomes we hope for. These approaches also deserve rigorous evaluation.
We should be willing to examine evidence that challenges our own recommendations. If research demonstrates that a particular strategy is ineffective, we should acknowledge that finding. If stronger evidence supports an approach we previously questioned, we should be prepared to reconsider our position. Scientific integrity should never depend on whether the findings support the position we already hold. That is the standard we believe all advocates, researchers, educators, policymakers, and technology companies should strive to meet. We also believe that disagreements about research interpretation should be approached with intellectual humility. Reasonable people can examine the same incomplete body of evidence and reach different conclusions about the most appropriate response, particularly when decisions involve competing risks, rights, and values. Such disagreements should encourage more thoughtful discussion rather than accusations that those holding a different position are indifferent to children’s safety, which is often a fallback strategy that we see “some” will use!
Parents, Caregivers, and Educators Deserve More Than a Selective Narrative
Parents and caregivers face difficult decisions about when and how to introduce technology into their children’s lives. They must consider their child’s maturity, developmental needs, family circumstances, educational requirements, social relationships, and individual vulnerabilities. These decisions are rarely simple, and we do not believe parents and caregivers are well served when complex scientific questions are reduced to emotionally compelling narratives suggesting that one parenting philosophy or policy approach is universally correct. We also do not believe they benefit when legitimate concerns about online harms are dismissed simply because some research identifies weak average associations between general technology use and mental health. Both approaches risk oversimplifying the evidence. Parents deserve access to research identifying potential harms, research documenting potential benefits, research questioning widely repeated assumptions, and research explaining the limitations and uncertainties that remain.
Educators deserve the same consideration when developing technology policies and digital literacy programs, particularly when those decisions affect students with different learning needs, accessibility requirements, and developmental circumstances. Policymakers should also be expected to consider the strongest available evidence when proposing legislation that may affect millions of young people. We recognize that reasonable people may examine the same incomplete evidence and reach different conclusions about the most appropriate response. That does not necessarily mean one side is ignoring science. Policy decisions also involve judgments about risk, rights, values, and acceptable trade offs. However, those judgments should be transparent and should not be presented as settled scientific conclusions when important questions remain unanswered. We believe there is an important difference between using research to inform a position and selecting research to defend a position that has already been established. Good science should inform the narrative. The narrative should not determine which science gets shared.
Our Invitation: Read the Research for Yourself
Rather than asking parents, caregivers, educators, and policymakers to simply accept our interpretation of the evidence, we have provided links to the original research so readers can examine the findings themselves. Our collection of 47 studies and reports includes information about research methodology, participants, study periods, key findings, limitations, and practical considerations. We encourage readers to approach this collection with the same critical thinking they should apply to any scientific research, including studies supporting our own position. We also encourage readers to examine systematic reviews and research documenting meaningful online harms. Our collection is not exhaustive, it is not intended to establish scientific consensus, and it is not an argument that technology is harmless. Rather, it is an invitation to consider credible findings that may not fit neatly into the narratives currently receiving the greatest public attention.
We believe that when parents, caregivers, and educators are given access to a broader range of credible evidence, they are better positioned to ask informed questions and make decisions appropriate to their circumstances. Our goal is not to replace one narrative with another, but to encourage a more complete and thoughtful conversation about youth, technology, and social media. We also hope that this collection encourages fellow advocates, researchers, and policymakers to examine the evidence that may challenge their own assumptions, just as we must remain willing to examine research that challenges ours. If we expect parents and caregivers to make evidence based decisions, then we should be prepared to provide them with access to the strongest available evidence, even when that evidence complicates the message we may wish to communicate.
The Other Side of the Story Deserves to Be Heard, but So Does the Responsibility to Prove Our Solutions Work
Here at The White Hatter, we have always believed in Facts Not Fear, Facts Not Emotions, Enlighten Not Frighten, and Know Tech Not No Tech. These are more than slogans. They reflect our belief that parents, caregivers, educators, and young people are best served when complex issues are approached through credible evidence, practical education, critical thinking, and informed decision making. We do not believe that recognizing credible research challenging sweeping claims about technology somehow diminishes the experiences of young people who have suffered genuine online harms. Both realities deserve recognition. We also do not believe that questioning the effectiveness of a proposed intervention means opposing efforts to protect children. Rather, it means taking those efforts seriously enough to ask whether they will actually work.
The central issue in today’s debate about youth, technology, and social media is not whether online harms exist, they do, and some can be devastating. We are also not questioning whether we should take action to protect children, because we absolutely should. The more important questions are whether we are considering the strongest available scientific evidence, whether we are accurately distinguishing between what research has established and what remains uncertain, and whether the interventions being proposed are capable of achieving their intended outcomes. We should not have to choose between protecting children and respecting scientific evidence, because those objectives should reinforce one another. We should also recognize that the risks associated with technology are not distributed equally, that children develop at different rates, and that no single intervention is likely to address every challenge associated with their digital lives.
For some families, delaying access to particular technologies may be appropriate. For others, carefully supported and age/developmentally appropriate access may provide meaningful opportunities for learning, connection, and developing digital competence. Schools may reasonably establish boundaries around personal technology use to support learning, while governments may reasonably introduce stronger safeguards to address foreseeable risks. Technology companies should be held accountable for designing safer environments and responding meaningfully to harms associated with their services. However, regardless of which interventions we support, we should be prepared to evaluate their effectiveness. If an intervention works, we should build upon it. If the results are mixed, we should investigate why. If it does not work, we should have the intellectual honesty to change course. If it produces unintended harm, we should be willing to acknowledge that as well. Importantly, we must apply this same standard to the approaches we personally advocate for, not just those we question.
Our children deserve more than policies driven by fear, public pressure, or even the best of intentions. They deserve protective measures informed by credible evidence, implemented thoughtfully, and evaluated independently. That is why we compiled these 47 research studies and reports. Not to prove that technology is harmless, not to dismiss those advocating for stronger restrictions, not to defend technology companies, and certainly not to minimize the experiences of children and families who have suffered online. We compiled them because a more complete understanding of the evidence can lead to better questions, better decisions, and ultimately better outcomes for young people. We believe that the science should shape the story, not the other way around, and that meaningful child protection requires more than identifying risks or introducing restrictions. It requires the willingness to determine whether our actions are producing the outcomes we intended.
When it comes to protecting children, the ultimate measure of success should never be how many restrictions we introduce, how many policies we pass, or how many people we persuade to agree with us. The ultimate measure of success should be whether our children are actually safer, healthier, better supported, and better prepared to navigate their onlife world. Because when all is said and done, the evidence that a problem exists is not the same as the evidence that a particular solution will work. That is a distinction we believe every parent, caregiver, educator, advocate, and policymaker should understand, and one that should guide what we do next.
Read Our Research Review
Technology, Social Media, and Youth Mental Health: A Chronological Review of 47 Research Studies and Reports (2026–2018)
Our annotated research guide brings together studies and reports examining technology use, social media, adolescent development, educational screen use, and mental health. It includes research findings, methodological considerations, limitations, and links to original sources. We encourage readers to explore the evidence, question our interpretations, and consider findings that both support and challenge their existing beliefs.
Digital Food For Thought
The White Hatter
Facts Not Fear, Facts Not Emotions, Enlighten Not Frighten, Know Tech Not No Tech
