Research and reporting on EdTech's impact on how children learn, focus, and develop. This conversation is happening nationally and globally. It should be happening here too.
The evidence that children learn to read better on paper than on screens is consistent, large-scale, and growing.
A meta-analysis of 54 studies (171,000+ participants, 2000-2017) found a consistent advantage for reading on paper over screens for informational texts. The advantage has grown, not diminished, as digital natives have come of age.
Delgado et al., Educational Research Review (2018) →In a study of 1,139 Norwegian 5th graders, children scored significantly higher on reading comprehension tests taken on paper than on identical tests taken digitally. The effect was largest for high-performing girls, despite near-universal access to digital devices at home.
Støle, Mangen & Schwippert, Computers & Education (2020) →A meta-analysis of 40 studies (469,564 participants) found that leisure print reading produces medium-size gains in reading comprehension, while leisure digital reading produces only a small effect. For primary and middle school students, the digital reading effect is negative.
Altamura, Vargas & Salmerón, Review of Educational Research (2023) →Digital devices in classrooms raise questions about attention and how students process information.
A 2024 OECD analysis found that 30% of students across OECD countries reported digital device distractions in every or most math lessons, with 59% distracted in at least some lessons.
OECD PISA 2022 Results, Volume II →EEG research shows that handwriting activates broader neural networks involved in memory and learning compared to typing, with stronger effects observed in younger children still developing literacy.
Van der Meer & Van der Weel, Frontiers in Psychology (2020) →Research with school-age children connects daily screen time to measurable effects on physical and mental health.
A systematic review and dose-response meta-analysis of 45 studies (335,524 participants, mean age 9.3 years) found that each additional hour of daily digital screen time was associated with 21% higher odds of myopia. Risk roughly doubled at 4 hours of daily screen time. Authors identified a potential safety threshold below 1 hour per day.
Ha et al., JAMA Network Open (2025) →In the US Adolescent Brain Cognitive Development Study, 9,538 children aged 9 to 10 were followed for two years. Each additional daily hour of screen time predicted higher depressive, conduct, and ADHD symptoms two years later, after controlling for baseline mental health, sleep, and physical activity. Effects are small at the individual level and meaningful at the scale of a student body.
Nagata et al., BMC Public Health (2024) →Recent research is beginning to clarify how generative AI tools affect student learning and how they treat different student populations when deployed in educational settings.
In a randomized controlled trial with nearly 1,000 high school students, those who used a ChatGPT-style AI tutor during practice scored 17% lower on subsequent unassisted exams than students who had no AI access. A second AI tutor with teacher-designed guardrails (giving hints rather than answers) avoided the negative effect. Analysis showed students used the unrestricted AI as a crutch, asking for and copying answers. Students did not perceive any reduction in their learning even when they performed worse.
Bastani et al., PNAS (2025) →Stanford researchers tested four widely used large language models on writing feedback for 600 eighth-grade essays. With prompts varying by student attributes, all four models produced systematically different feedback for the same essays. Students presumed to be Black, Hispanic, Asian, English language learners, or to have a learning disability received less substantive critique and more praise. Students presumed to be White, male, or motivated received more direct engagement with reasoning. Effects appeared even when only student names were provided. The authors note these patterns appear in LLM-based feedback systems currently being piloted in US school districts.
Tan, Phalen & Demszky, LAK 2026 →Software intended to protect students can create its own privacy concerns.
GoGuardian, used by BISD to monitor student activity, was flagged by the Electronic Frontier Foundation for collecting browsing data, search history, and YouTube activity, even outside school hours. The software designed to keep children safe raised its own privacy concerns.
EFF: How GoGuardian Invades Student Privacy →A March 2025 class action lawsuit alleges that Instructure, the parent company of the Canvas learning management system, collects messages, search activity, test results, essays, and photos from student users. The complaint asserts the collected data is used to "build dynamic, robust, and intimate dossiers of children."
Robinson+Cole Data Privacy Insider (2025) →This is not a fringe conversation. From Kansas to Sweden, school systems are reassessing 1:1 device policies and asking whether the evidence supports what they have been doing. Here is some of the recent coverage.
Schools in North Carolina, Virginia, Maryland, Michigan, and Kansas are rethinking 1:1 laptop policies as studies show device access has coincided with stagnant or declining test scores.
McPherson Middle School in Kansas collected all 480 student Chromebooks, returning to paper-based learning. The principal, named Kansas' 2025 middle school Principal of the Year, said the devices were too great a distraction.
California's third-largest district is pulling take-home laptops from 40,000 elementary students, shifting to classroom-only use to increase instructional time and reduce unsupervised screen exposure.
After years of declining literacy scores, Sweden is investing over 100 million euros to restore printed textbooks, restrict preschool screen use, and implement a national phone collection policy in schools.
A 2025 government study found that three-quarters of Dutch high schools reported better student concentration after banning phones, tablets, and smartwatches from classrooms.
Mesick Elementary pulled Chromebooks and iPads from K-5 classrooms after 65% of third through fifth graders scored "partially proficient" or "not proficient" in English Language Arts. The district allocated $30,000 for books and hands-on materials.