There's a conflation in the public conversation around screen time between access to devices and the choice to use the devices that students have access to. These are related but distinct issues, and they should be considered differently. 

I hope to convince you that when it comes to deciding whether to use a screen in a classroom, we have to defer—in the interest of the greatest good for the greatest number of students—to those closest to the students: their teachers. But I also want to argue that teacher judgment only works inside a frame, and that frame is the legitimate work of policy. Getting the line between the two right is what this blog is about.

A note on where I'm coming from: I helped found and now help lead Newsela, a company whose products live primarily on screens. Newsela also supports plenty of printing, and I take seriously the research suggesting that, in some cases, paper is better than digital. (Transparently, we have Newsela research suggesting that digital use is associated with better student outcomes.) The argument that follows is the one I'd make regardless of where I worked.

• Where are we now?
• What policy should do
• What policy should not do
• What this looks like in practice: A writing example
• The line between policy and instructional decisions
• How are district leaders navigating this line?
• Follow the conversation about AI in the classroom

[Where are we now?](id-where)

The current debate didn't come from nowhere. During the COVID years, districts were flush with one-time funding and racing to make remote learning work. A lot of software was purchased without vetting. A lot of tools entered classrooms without a framework for how they'd be used or what they'd replace. When the federal funding dried up, many tools stayed, but the cohesive vision for how they support student learning didn't catch up. 

AI has arrived in this environment, and is arriving faster than most districts can govern.

So when parents say things feel out of control, or when legislators say they can't get straight answers about what's happening with their kids and their data, they are responding to something real. The skeptical reaction to screens in schools is not irrational. It's an accountability vacuum being filled, late, by people too far away from the classroom.

The argument is not "trust teachers and back off." Rather, there is real work for policy to do, and there is real work that policy cannot do. Conflating the two is how we end up with bad outcomes at both ends.

We’ve been here before, and we’ve learned from it

In a book published in 2001, Oversold and Underused, Larry Cuban argued that schools absorb technology without changing the underlying instructional model. The technology gets domesticated. It becomes whatever the school was already doing, just on a screen. 

The implication? If your instruction was incoherent before the Chromebooks arrived, the Chromebooks will be incoherent. If your writing curriculum lacked rigor, the laptop won't add it.

The good news is that there is a history of making decisions about new technology and helpful frameworks that have been around for at least a couple of decades. 

On a recent panel I hosted with administrators from Massachusetts, California, Missouri, and Georgia, a San Diego administrator referenced the SAMR model for thinking about technology use in schools. The SAMR model gives us a structure to thoughtfully consider the benefits and tradeoffs of our use of technology.

For example, using a screen simply to substitute paper reading materials with a digital article (“Substitution” in the SAMR model), or perhaps “Augment” the article by allowing for a larger font size and read-aloud mode, can enhance instruction, but it doesn’t necessarily transform it. Transformation comes into play when technology allows for significant task redesign (e.g., auto-grading student responses in real-time for in-the-moment teacher intervention) or allows for the creation of new tasks, previously inconceivable (e.g., giving each student rubric-aligned feedback as they practice and revise their writing—for every student, at the same time). 

So, what can we learn from previous waves of technology adoption in classrooms?

Don't lead with the technology. Start with what you are trying to accomplish for students. Design your policy and models first, then ask what tools will serve those models while fitting within the policy. That's the reasonable framework Larry Cuban was calling for 25 years ago when he observed that schools were doing just the opposite. 

[What policy should do](id-policy)

There is a meaningful floor and perimeter that policy should set, and I think most of it should be set at the district level with state-level minimums in specific areas. Here is what I'd actually endorse:

Data privacy and vendor vetting 

Districts, not individual teachers, should decide which platforms collect what data on students, how long that data is retained, and who it's shared with. This is not a pedagogical question and should not be a teacher-by-teacher decision.

Age-appropriate access

The developmental evidence for screen use in the earliest grades is different from the evidence for older students. A policy cap on K–2 screen time is reasonable and warranted.

Phones as a separate category 

Cellphones in schools are a different issue from instructional technology and should be treated as such. Although phones have screens, the policies being enacted in nearly every state and district to severely limit or ban phone use during instructional time have little to do with technology for teaching and learning. 

The overwhelming use of phones in schools has been for personal, consumer-style behavior. No serious proponent of intentional screen use includes cellphones in that argument. Bell-to-bell phone-free instructional time is a legitimate exercise of administrative authority. (I'll add that in my own household, we use screen limit technology to govern my son's use of consumer apps like Instagram and YouTube: tools that incentivize use that’s anything but educational. The instinct to wall those off is the right one.)

Transparency to families 

Districts should publish what tools they use, what those tools do with student data, how AI is being used in instruction or feedback, and how families can ask questions or opt out. 

The "can't get answers" problem is fixable, and fixing it would do more to lower the temperature of this debate than any single legislative intervention.

AI-specific guardrails 

Products must disclose when AI is used in instruction or feedback. 

Limits on AI making consequential decisions about students’ placement, grading, and identification for services belong in policy.

Teacher preparation as a precondition 

Teacher judgment is only as good as teacher preparation. Districts have an obligation to ensure that the teachers being trusted to make these decisions have actually been trained to make them. This is the part of the deal that gets skipped most often but matters most.

This is a real list; it’s not simply "trust teachers." It is a serious policy program. And nothing on this list legislates the instructional decision itself, which is the next part of the argument.

[What policy should not do](id-boundaries)

In the 2021 second edition of Why Don't Students Like School?, cognitive psychology professor Daniel Willingham offers a question that I think should be at the center of any decision about classroom technology: "When new technology replaces old, something is sacrificed. Am I clear on what that is, and am I comfortable with it?" 

Willingham’s example is the chalkboard versus the overhead projector. The chalkboard preserves everything the teacher wrote earlier in the lesson, still visible at the back of the room. The projector shows one thing at a time. Does that matter? How much? That's up to the teacher’s judgement.

Willingham's question is the right one, and it can only be answered by those close to the student. It depends on three dynamic variables that can't be effectively legislated from beyond the classroom:

• The learner and their context. Who is the learner, and what do they bring to this assignment by way of skills and knowledge?
• The assignment and its purpose. What knowledge or skill is most important for them to learn, and how is this assignment aligned to that goal?

The tradeoff with screens. What does research and my own experience predict my students will sacrifice—and gain—by my allowing a screen-enabled device versus a comparable method?

The role of administrators is to craft policies—with teacher and student input—that frame this decision with the right questions and the right data. Their role is not to make the decision for the teacher, and certainly not to have it made in the statehouse.

[What this looks like in practice: A writing example](id-writing)

Writing is a good example of how research can help us understand effective practices across modalities—what each is good for, why, and for whom. 

When teachers decide whether to assign an activity that requires handwriting vs. typing, different student groups experience different tradeoffs related to comprehension, information retention, editing and revising, screen distractions, and teacher insights. 

Writing by hand can be helpful for planning long-form writing as well as encoding information. But when it comes to revising work, it’s easier to do this digitally. Neurodivergent students or students who physically struggle with handwriting (e.g., students with dyslexia or dysgraphia) may benefit from a screen-enabled device used to capture their ideas through accommodations like speech-to-text. And multilingual learners or students reading below grade level may benefit from some of the additional scaffolding that’s difficult to provide outside of digital devices.

If we want students to take their time to develop ideas thoughtfully and make a plan to write their essay before diving in mid-thought, we should encourage them to start on paper. If we want them to take advantage of feedback (from any source—a peer, their teacher, AI), digital writing will be best for this.

Decisions need to be made on an individual basis—for students, assignment by assignment—as to whether a screen will help or harm. Some assignments will take longer than others. Some students will need screens for the whole assignment; others won't need them at all. Setting a timer, or worse, legislating limits through technology like screen locks, is counterproductive to the actual goal.

[The line between policy and instructional decisions](id-line)

What I'm arguing for is that policy defines the boundaries. Districts vet vendors and protect student data. States and districts keep phones out of instructional time. Families know what tools are being used and how AI is involved. Teachers are prepared to make the decisions they're being asked to make.

Inside that frame, the decision about whether to use a screen for a given assignment belongs to the teacher—because that decision depends on the learner, the assignment, and the tradeoff, and these variables can only be read accurately by someone in the room.

The further we get from the classroom, the less in focus our view becomes, and the less wherewithal we bring to the question of whether and why to use or forgo a screen in today's lesson. This is the case against legislating instructional screen use. It is not a case against governance. It is a case for governing at the level where governance can actually work.

We’ve spent the last several years lurching between under-governance and over-governance—districts that didn't vet what they bought, and now legislatures trying to reach into classrooms to fix it. Neither extreme serves students. The work in front of us is to get the balance right. That's not a slogan, and it doesn't fit on a yard sign. But it's the actual job.

[How are district leaders navigating this line?](id-districts)

Every district’s needs are different, as are each student’s needs. For an overview on how districts across the U.S. are building their screen time policies—and how those translate to classroom instruction—check out this panel discussion featuring leaders from San Diego USD, DeKalb County Schools, Blue Springs School District, and Lexington Public Schools.

“We've learned a balanced approach [to screen time policies] also entails trusting and giving our teachers professional learning so that they can make great professional decisions versus compliance. When we're saying across the board that all of the students in your class should do  times a week, that really takes out [teachers’] professional judgment as well as goes against our purpose and that technology should be intentional.” – Director of Elementary Education, Blue Springs School District (MO)

[Follow the conversation about AI in the classroom](id-podcast)

When it comes to the screen time debate, approaches to AI in the classroom are a major component driving questions, practices, and policies. Stay up to date with the latest conversations in education with a podcast from Chief Academic Officer & Co-Founder Dan Cogan-Drew: AI in the Classroom - Daily.