Teaching with Technology in the Classroom: What Works and What Doesn't

Every decade or so, a new technology enters education with the promise that it will transform learning — that it will close achievement gaps, engage disengaged students, and revolutionize what teachers can accomplish. Most of these promises have failed to materialize at scale.

Television was going to transform education in the 1960s. Computer labs were going to do it in the 1980s. Interactive whiteboards were going to do it in the 2000s. 1:1 device programs were going to do it in the 2010s. And now AI is making similar promises.

None of this is to say technology has no value in classrooms — it has significant value, when used in specific ways. But the research tells a more complicated story than most EdTech marketing suggests, and teachers benefit from understanding that story before deciding what to adopt and how.

What the Research Actually Shows

The research on educational technology is large, highly variable in quality, and yields a simple pattern: technology works when it's in service of sound pedagogy, and fails when it's expected to substitute for it.

Studies that show technology producing gains typically share a few features: the technology provides targeted, adaptive practice at the right level (adaptive math software, for example); students receive immediate feedback that they can act on; teachers use technology as one tool within a broader instructional system rather than as the primary instructional mode; and teachers received training in using the specific technology effectively.

Studies that show technology producing neutral or negative effects typically show: replacing teacher-led instruction with screen time; passive consumption (watching videos) substituting for active learning; off-task use that's difficult to monitor; and implementation without training or clear pedagogical rationale.

The effect size for educational technology overall is small to moderate — meaningful but not transformative. The effect size for great teaching is much larger. Technology amplifies what's already happening; it doesn't substitute for it.

Where Technology Genuinely Helps

Adaptive practice platforms: Software that adjusts difficulty based on student performance and provides immediate feedback produces measurable gains in math and reading fluency. Khan Academy, IXL, and similar platforms do this reasonably well for procedural skills. The benefit is that students practice at their current level rather than a whole-class level that's too easy or too hard.

Research and information access: Giving students access to a wider range of sources than any physical library could contain is a genuine capability expansion. The challenge is that the same access that enables research also enables distraction, and students need explicit instruction in evaluating sources online.

Writing and revision tools: Word processing makes revision genuinely easier than handwriting — students can rearrange, delete, and insert without starting over. Research shows students revise more when writing on computers. The caveat: this benefit depends on students actually revising, which requires explicit instruction and structured revision activities rather than just "make it better."

Presentation and creation tools: Students who create multimedia artifacts — videos, podcasts, visual explanations — often demonstrate understanding in ways that traditional writing doesn't capture. Creation projects can be highly engaging and authentically assess complex understanding.

Formative assessment tools: Polling tools (Mentimeter, Poll Everywhere, Kahoot for review) give teachers real-time information about student understanding in ways that raise-your-hand-if-you-understand doesn't. The ability to get an anonymous, simultaneous response from every student is a genuine instructional improvement.

Where Technology Doesn't Help (or Makes Things Worse)

Replacing instruction: A video of someone explaining photosynthesis is not the same as a teacher explaining it interactively, fielding questions, checking understanding, and adjusting based on what students need. Video-based instruction (flipped classrooms, recorded lectures) works for highly motivated, self-directed learners; for typical K-12 students, it produces substantially lower outcomes than live instruction.

Passive consumption: Watching a documentary, a YouTube explainer, or a recorded lesson does not typically produce durable learning, for the same reasons that listening to a lecture without active processing doesn't. Screen time is not the same as engaged learning time.

Devices without managed use: A 1:1 device program without clear structures for on-task use, without teachers who have the tools and authority to monitor and redirect, and without student culture around responsible use often produces more distraction than learning. The device opens every possible off-task activity while also opening instructional content.

Novelty for its own sake: The EdTech industry produces an enormous volume of products, many of which have no evidence base and are adopted because they're new and interesting. Teachers who adopt tools because they're exciting rather than because they serve a specific instructional purpose often find that the excitement fades quickly and the learning gains don't materialize.

How to Decide What to Use

A few questions worth asking before adopting a new tool:

What specific instructional problem does this solve? If you can't name the problem, the tool probably isn't the answer.

What will students be doing with it? Actively creating, practicing, and responding are higher-value uses than passively watching or reading.

How will you manage off-task use? If you don't have a clear answer, off-task use will happen.

What does the evidence say? Many EdTech products cite research studies prominently; check whether those studies are independent, rigorous, and actually about the product you're considering rather than the category of product.

What happens when it doesn't work? Technology fails. Networks go down. Accounts don't work. Apps update and break. Have a plan for what students do when the technology isn't available.

AI in the Classroom: The Current Question

Generative AI tools — ChatGPT, Claude, and their successors — are the current disruptive technology in education. The honest answer about their educational value is: complicated, context-dependent, and rapidly evolving.

AI can usefully generate content for teacher preparation: lesson plans, discussion questions, practice problems, writing prompts, differentiated versions of materials. LessonDraft uses AI specifically for this purpose — accelerating the preparation work that takes teachers significant time without changing what happens between teachers and students.

For student use, the questions are harder. AI can help students brainstorm, get feedback on drafts, and explain concepts. It can also do students' work for them in ways that undermine the learning the assignment was designed to produce. Schools and teachers are still working out when and how AI use serves learning and when it subverts it — and there's no settled consensus yet.

Your Next Step

Identify one technology tool you use regularly. Be honest about whether it's producing learning gains or just using time. If you're not sure, try removing it for two weeks and notice what changes.