Week 12 — How Does My Feed Know Me?
The Algorithmic Echo — Part 1
This week introduces one of the most important concepts in modern media literacy: the algorithm. Students learn that the content they see online is not random — it's selected by a set of rules designed to show them content they're likely to engage with — based on what they've clicked, watched, and liked before. Every like, watch, and scroll teaches the algorithm more about them. The feed is not a mirror of the world; it's a mirror of what the platform thinks will keep them watching.
Key Vocabulary
| Term | Definition |
|---|---|
| Algorithm | A set of rules that a platform uses to decide what content to show you, based on your past behavior |
| Feed | The stream of posts, videos, or articles a platform shows you — curated by an algorithm, not randomly selected |
| Engagement signal | Any action you take (like, comment, share, watch time) that tells the algorithm what to show you more of |
| Personalization | When a platform customizes what you see based on your behavior, so that two people using the same app see different content |
Imagine a librarian who watches everything you read and only gives you more of the same kind of book. That's kind of what an algorithm does — it watches what you click, like, and watch, and then shows you more of that. It's not random. It's not showing you everything. It's showing you what it thinks will keep you interested.
Connection
Units 1–3 taught students to see how media is built, paid for, and verified. This week asks a new question: who decides what you see in the first place? Students discover that algorithms curate their feeds based on engagement, not accuracy or value. Next week they explore what happens when that curation goes too far: filter bubbles and echo chambers.
From Weeks 5-6: Students learned about the attention economy and clickbait. Now they see the system behind it: algorithms use engagement signals (the same clicks and watch-time they studied) to decide what to show next. The attention economy is powered by algorithmic curation.
From Week 2: "Who made this and why?" now includes a new question: "Who chose to show me this, and why?"
Teacher Preparation
Prepare the following:
- A simple analogy for "algorithm" (the lesson uses a librarian analogy — see below)
- A smartphone or tablet to demonstrate how recommendations work (optional — you can also describe it or draw it)
- Paper and markers for the "Recommendation Game" activity
- Two different YouTube search/home pages (if possible) to show how different accounts see different things — or describe this scenario verbally
Note: You don't need the student to be on social media for this lesson. Most kids understand the concept of "recommended videos" from YouTube, "For You" on TikTok, or "suggested" content in games.
Write 10 content topics on index cards (funny cats, soccer, cooking, science, scary stories, etc.). That's all you need for the Recommendation Game. No device required — the paper simulation teaches the concept perfectly.
Many kids assume the internet shows everyone the same thing. The realization that their feed is personalized — and shaped by engagement signals rather than by what's most accurate or most valuable to the viewer — is often a genuine surprise. Let that moment land. Don't rush past it.
Guided Session 1
What Is an Algorithm?
Learning Goal
Students can explain what an algorithm is (a set of rules that decides what to show you) and understand that platforms use algorithms to keep users watching.
Activities
-
The Librarian Analogy — Ask: "Imagine a librarian who watches everything you read. Every time you pick up a dinosaur book, she writes it down. Every time you put a space book back on the shelf after one page, she writes that down too. After a few weeks, she stops suggesting space books and ONLY hands you dinosaur books. She's not showing you the best books — she's showing you the books she thinks you'll keep reading."
"That librarian is an algorithm. Except on the internet, she works instantly, and she's watching millions of people at the same time."
-
How the Feed is Built — Explain the basic mechanics:
- You watch a video all the way through → the algorithm thinks: "they liked that — show more like it"
- You skip something after two seconds → the algorithm thinks: "they didn't like that — show less like it"
- You like, comment, or share something → the algorithm thinks: "strong signal! Show MUCH more like it"
- You search for something → the algorithm adds that topic to your profile
The result: your feed becomes more and more specific over time. It's not showing you the world — it's showing you a version of the world that the platform thinks will keep you engaged.
-
The Recommendation Game — On paper, simulate an algorithm. Create cards with different types of content: "Funny cat video," "Science documentary," "Soccer highlights," "Scary story," "Cooking tutorial." Give the student 10 cards and have them "watch" 5 by picking the ones they'd click on. Then remove the unclicked cards, double the topics they chose, and present a new round. And again. After 3 rounds, look at the feed: "It's all the same type of content now. How did that happen?"
-
Anchor Concept — Introduce the third core concept: Algorithms shape what you see. Explain: "Many recommendation systems are designed to keep you engaged — meaning they tend to show you content you're likely to click on, watch, or interact with. This often means emotional or exciting content gets promoted, while calmer or more accurate content may get less attention. This is a simplified model — platforms are complex, and some are working to promote quality content too — but it's a useful way to start understanding how your feed is built. The key idea: the algorithm doesn't choose what's true or important. It chooses what's likely to keep you watching."
Epistemic guardrail: Algorithms are systems — not magic, not automatically bad. They're designed by people to accomplish a goal (usually engagement). Some algorithmic recommendations are genuinely helpful: they surface music you love, tutorials you need, or news you care about. The issue isn't that algorithms exist — it's that their goal (keep watching) and your goal (be well-informed, make good choices) aren't always the same. Understanding the system gives you more control, not less.
-
Media Checkpoint Connection — Link to The Media Checkpoint. Every question in the checkpoint still applies — but now add a new layer. Before you even start analyzing a piece of media, ask: Why am I seeing this in the first place? The algorithm chose to show it to you. That's a construction choice too — made by software instead of a person.
Reflection Questions
- Before today, did you think about why you see certain videos or posts?
- How is your feed different from what a friend or parent might see?
- If the algorithm only shows you things you already like, what might you NEVER see?
Guided Session 2
The Recommendation Experiment
Learning Goal
Students can describe how interacting with content (liking, watching, searching) changes what the algorithm shows next, and begin to see their feed as a curated, not neutral, experience.
Activities
-
Before and After — If you have a device available, look at the home page or "recommended" section. Take a screenshot or write down the top 5 suggestions. Then deliberately watch or interact with something very different from what's usually shown (if the feed is all gaming, watch a cooking video all the way through). Check the recommendations again later. Did anything change? (Even a small shift demonstrates the algorithm responding.)
If a device isn't available, describe this experiment verbally and discuss what the student would expect to happen.
-
Two Feeds, One World — Discuss: "If Person A watches funny animal videos all day, and Person B watches news about science, their feeds will look completely different. They're using the same app, living in the same city — but they see a different internet." Ask: "Is either feed showing 'the truth'? Or are both showing a piece of the truth selected by the algorithm?"
-
The Engagement Trap — Explain why emotional content gets boosted: "If a calm, factual video gets watched for 2 minutes but an outrageous, dramatic video gets watched for 7 minutes plus 50 comments, which one will the algorithm promote? The dramatic one — because it generates more engagement. This doesn't always happen — platforms are working on ways to promote quality content too — but the general pattern holds: engagement signals are powerful, and emotional content tends to generate more of them."
-
Draw Your Algorithm — Have the student draw or diagram their "ideal feed" vs. what an algorithm would probably build for them. The ideal might include a variety of topics; the algorithm's version would be narrower and more repetitive. Discuss the gap.
Reflection Questions
- If the algorithm had its way, what would your feed look like in a year?
- Does the algorithm show you what you need to know or what you want to watch?
- How could you "train" an algorithm to show you a better, more balanced feed?
Independent Session
Algorithm Journal
Instruction
For the next few days (starting today), keep an Algorithm Journal. Each time you use a device with recommendations (YouTube, a game with suggested content, an app), write down:
- What was recommended — the first 3–5 things shown to you
- Why you think the algorithm suggested them — what have you watched, liked, or searched recently that might explain it?
- What's MISSING — what topics, perspectives, or types of content are you probably NOT seeing because of how the algorithm has learned your preferences?
Today, fill out at least one journal entry. If possible, continue updating the journal over the week. At the end of the week, review your entries and write a one-sentence summary: "My algorithm seems to think I am someone who ________."
Skills Reinforced
- Observing algorithmic curation in real time
- Connecting personal behavior (clicks, watches, likes) to algorithmic output
- Identifying what a personalized feed excludes
Setup
Provide a journal or notebook with three columns: "What Was Recommended," "Why I Think It Was Chosen," "What's Missing." The student can make entries throughout the week. For today's session, have them make their first entry using whatever device and app is most frequently used. Set a timer for 20 minutes.
Quick Check
After this week's sessions, the student should be able to:
- Define algorithm: Explain what an algorithm does in their own words (no jargon — the librarian analogy is fine).
- Name the signals: List at least three engagement signals that teach the algorithm about them.
- Spot the pattern: Look at their own feed and explain why certain content appears there.
Caregiver Look-Fors
- The student uses the word "algorithm" naturally when talking about what they see online
- They connect their own actions (watching, liking) to what shows up in their feed
- They show genuine surprise at how personalized their feed is
- They ask questions like "Why am I seeing this?" or "What is the algorithm optimizing for?"
- The Recommendation Game produces a convincing narrowing effect
🎯 Takeaway
Big idea: The content you see online is selected by algorithms that respond to your behavior — not by someone choosing the best or most accurate content for you.
Remember: Your feed is shaped by your clicks, but it doesn't have to define what you think. You can always search, explore, and choose to look beyond what's recommended.
Younger Learner Adaptation (Ages 6–8)
- Lean heavily on the librarian analogy: It's concrete and memorable. Act it out with real books if possible.
- Simplify the Recommendation Game: Use 5 cards instead of 10, and run only 2 rounds.
- Skip the Algorithm Journal for now: Replace it with a single conversation: "What does YouTube always show you? Why?" That's the concept.
- Use a familiar app: If the student plays a game with recommended content, use that as the example.
Older Learner Extension (Ages 11–13)
- Compare platforms: How do YouTube, TikTok, and a news app differ in how they build feeds? Which is most aggressive about engagement?
- Data collection discussion: What data does the platform collect to build your profile? How do you feel about that?
- Helpful vs. unhelpful recommendations: Find three recommendations that were genuinely useful and three that felt like time-wasters. What's the difference? Can you tell when the algorithm is serving your goals vs. the platform's goals?
- Experiment: Over 3 days, deliberately interact with content outside their usual pattern and track how the feed changes.
Accessibility Options
- Draw the algorithm: Instead of writing the journal, draw what the algorithm "thinks" they like.
- Verbal journal entries: Record observations as voice memos instead of writing.
- Physical Recommendation Game: Use actual objects (toy figures, snack boxes, book covers) as content cards.
- Partner journal: Adult and student keep the Algorithm Journal together, comparing observations.