Students in Grades 6–12 live inside recommendation systems and have almost no language for them. Tell a class that algorithms shape what they see and you get either a shrug or a conspiracy theory. Neither is useful. What changes their thinking is doing something that makes the sorting visible, then having to explain what they saw. These five activities are low on preparation, run in a single period each, and give students evidence rather than a warning.
Why lecture fails on this topic
Filter bubbles are hard to teach because the effect is invisible from the inside. A student's feed looks like the world, not like a selection, and no amount of explaining changes that impression. The other obstacle is defensiveness. Students hear "you are being manipulated" as an insult to their judgment, and they stop listening. Every activity below is built to avoid both traps: the student generates the observation, and the framing is about how the system works rather than about how gullible the user is.
Five activities that make the sorting visible
- The parallel feed comparison. Four students search the same neutral term on their own devices and compare the first five results or suggestions. Nobody shares anything personal, only the ordering. Purpose: personalization stops being a claim and becomes an observation they made themselves.
- Signal audit. Students list every action they take on a platform that a system could read as a preference: watch time, rewatch, pause, follow, mute, skip. Rank them by how much each probably counts. Purpose: builds the idea that the input is behavior, not stated opinion.
- Design the algorithm. In groups, students write the ranking rules for a fictional platform whose goal is maximum time on site, then rewrite them for a platform whose goal is informed users. Compare the two rule sets. Purpose: makes objective functions concrete, and shows that the design is a choice rather than a fact of nature.
- Same story, three framings. Give one event described three ways by three sources. Students identify what each version foregrounds and who it is written for. Purpose: separates the platform question from the source question, which students routinely conflate.
- Deliberate diversification. Students choose one topic they hold an opinion on, find the strongest version of a view they disagree with, and summarize it fairly in five sentences. Purpose: rehearses the actual skill, which is seeking out what the feed will not deliver.
Activity three tends to run long and is worth planning properly, since the group rule sets are the assessment evidence for the whole sequence. The participation and manipulation strand in Social Media & Digital Platforms – Participation, Influence, and Manipulation supplies the case material and structured tasks so you are not inventing platform scenarios on a Sunday evening.
Handling the AI question when it comes up
It will come up, usually during activity two, and usually as "is the algorithm an AI." Answer it plainly. Recommendation systems learn patterns from behavior data, and that is the same family of technique behind the writing and image tools students already use. The productive follow-up is not whether this is good or bad but what the system is optimizing for and who chose that target. Keep the conversation off specific brands, because naming products turns a structural discussion into a loyalty argument between students. The commercial side is where this gets sharpest anyway, since sponsored content is where recommendation and persuasion meet, and the classroom material in Influencers and Social Media – Ethics Unit: Worksheets, Projects & Slides (Grades 6–9) gives you disclosure and sponsorship scenarios that middle schoolers recognize immediately.
Landing it somewhere that matters
End the sequence on civic ground rather than personal habits. Ask what happens to a vote, a referendum, or a local decision when different groups of citizens see genuinely different information about it. That question turns a lesson about feeds into a lesson about how public opinion forms, and it is the point where older students start taking the topic seriously. Material on citizen participation and how proposals reach a decision, such as Direct Democracy Explained – Political Participation & Voting | Social Studies Resource | GRADE 5–9, works well as the bridge lesson if your school links digital literacy with social studies. The class you are aiming for is one where a student scrolls past something and says, without being asked, that they were shown it for a reason.
Related reading from the Teacher Hub
- Your First Lesson on How the Internet Works: What to Cover and How (AI & Digital Literacy, Grades 6–12)
- Teaching Digital Citizenship: Common Misconceptions and How to Fix Them (AI & Digital Literacy, Grades 6–12)
- How to Assess Data and Databases: Rubrics, Exit Tickets and Feedback (AI & Digital Literacy, Grades 6–12)


Comments
No comments yet — be the first to share your thoughts!
Leave a comment
Comments are reviewed before being published.
Thanks for your comment!
Your comment is being reviewed and will appear here shortly.