Teacher Pack

Everything you need for a teaching unit on “How do feeds & algorithms work?” — worksheets, class test, homework, parent letter and curriculum mapping. Designed for lower secondary (Years 6–9), subject media literacy / digital literacy. Free to use under CC BY 4.0 with attribution to “Webagentur Hochmeir e.U. (webhoch.com)”.

📚 Years 6–9 (lower secondary) ⏱ Double lesson 90 min · Weekly module 3 × 50 min · 5-week project 📱 Fully printable (Ctrl/Cmd + P)

How do I use this pack?

The teacher pack complements the main page in teacher mode with everything you need to print. It is built so that you can print it once and reuse it for years. The interactive filter-bubble simulator on the main page works beautifully as an opener on the projector — the class watches in real time how a feed narrows.

Core stance: mechanics, not morals

This pack explains how feeds and algorithms work — calmly and factually. It does not condemn any platform or any usage habit. Terms like “endless scrolling” or “comparison pressure” are treated mechanically and in an empowering way: we describe the mechanism behind them and what students can take into their own hands — not what they are “doing wrong”. Please introduce the topic in this tone.

Recommended sequence

  1. Preparation: First read through the main page in teacher mode. For each chapter it gives you learning goals, timing, method tips, quiz answers and “mini worksheet” tasks.
  2. Print the material: Print the worksheets that follow here (ideally one per student minus one, with one held in reserve). Print the parent letter as a class set.
  3. In the lesson: Follow the lesson overview. Show the filter-bubble simulator on the projector; the worksheets serve as an activity, not a test.
  4. Assessment: Optional class test at the end of the unit. Grade using the included rubric.

What's included?

Printing tips

Data & accounts — important note

This unit requires no social-media account. The simulator runs locally in the browser, collects no data and needs no sign-in. Every homework task has a variant solvable without a personal profile (e.g. comparing two public topic searches). No real profiles, real names or private feeds are shown in class.

Lesson overview & learning goals

The unit answers a single big question: Who actually decides what I see in my feed? From there it opens up ranking, the attention economy, filter bubbles, outrage amplification and healthy use.

Overarching learning goals

By the end of the unit, students can …

Competence matrix (cognitive demand levels)

LevelDemandExample from this unit
Level I — ReproductionName terms and signalsList the four engagement signals a ranking rewards
Level II — ApplicationExplain and apply connectionsShow in the simulator how a like shifts the next batch
Level III — Transfer / evaluationJudge, take a position“My feed contract”: justify your own usage rules

Lesson plan — double lesson (90 min)

TimePhaseContent & method
0–10Warm-upQuestion “Why does scrolling often not stop?” — collect guesses (board).
10–25Development 1The feed is built, not neutral; what an algorithm sorts (the waiter image).
25–45Development 2Filter-bubble simulator on the projector + Worksheet 1 (Be the algorithm).
45–60Development 3Attention as the currency + spotting ads (Worksheet 2).
60–80Development 4Filter bubble vs. echo chamber; why outrage is amplified (Worksheet 3).
80–90ConsolidationHealthy use — start “my feed contract”, quiz review.

In a single double lesson the selection is tight — pick 2 of the 4 worksheets. For the full depth, use the weekly module (3 × 50 min) or the 5-week project (see curriculum mapping).

Worksheets

Each worksheet suits about 15–20 minutes of individual or pair work. The answer key is in a fold-out box directly underneath — cut it off before printing the class set, or print double-sided (student side at the front, answer key at the back — do not hand to students).

Topic 1: Be the algorithm Name: ______________________ Class: ______________________ Date: ______________________

Worksheet 1 — Be the algorithm

Imagine you are the algorithm. Your job is not to show the “most important” posts, but the ones the person stays with longest.

1. Order the signals (4 points)

A feed ranking rewards some reactions more than others. Order the four signals by their strength (1 = strongest signal, 4 = weakest). Write the number in the circle:

  • ◯ Like (❤)
  • ◯ Dwell time (how long you stay on a post, 👁)
  • ◯ Share (➡ forward to friends)
  • ◯ Comment (💬)

2. Predict the next batch (3 points)

Someone likes three sports posts in a row. Which category will the feed probably show more of afterwards — and which less? Explain in 1–2 sentences:

3. Ranking is a prediction (3 points)

Explain in your own words: which question does the algorithm “ask” about each post before showing it high up?

4. Chronological or algorithmic? (2 points)

What is the difference between a chronological feed (newest first) and an algorithmic feed (ranked)? Name one advantage of each:

5. Reflection (2 points)

In the simulator the “diversity index” drops the more you feed one category. What surprises you about that? Explain briefly:

Total points: ___ / 14

🔑 Answer key for teachers — Worksheet 1

1. Order of strength: Share (1) → Comment (2) → Like (3) → Dwell time (4). Reasoning: actively forwarding and writing show more interest than a quick tap; dwell time is a weak but frequent signal. 1 point per correctly placed signal.

2. Afterwards more sports, less of the other categories. Accepted: the insight that the same reactions raise a category's weight and the next batch follows it.

3. In effect: “How likely is this person to react to this post (stay, like, comment, share)?” Ranking is therefore a prediction of expected engagement, not a judgement about quality or truth.

4. Chronological: predictable, you see everything in time order; nothing is “filtered out”. Algorithmic: shows first what is likely to grab you; can surface relevant things — but can also narrow.

5. Accepted: any well-reasoned answer (e.g. “that a few likes are enough to tip the feed”, “that I don't even notice how one-sided it gets”).

Topic 2: The ad detective Name: ______________________ Class: ______________________ Date: ______________________

Worksheet 2 — Attention is the currency

Many services cost no money. Still, they earn. This sheet shows how — and how you spot ads in the feed.

1. Why free? (3 points)

An app is free. So how does it make money? Explain the connection between your attention and the business model in 1–2 sentences:

2. The ad detective (3 points)

Here are invented feed posts. Tick which ones are a paid ad, and note the giveaway:

PostAd? Giveaway
“Sponsored · NewShoes — now −30%”☐ yes ☐ no _______________
A friend posts a holiday photo☐ yes ☐ no _______________
“Ad · Learn guitar in 7 days”☐ yes ☐ no _______________

3. The goal: stay long (2 points)

Why is it valuable for the service that you stay in the app as long as possible? Name the reason:

4. What belongs where? (2 points)

This site explains why advertising is an incentive. Which data is collected and how the ads are auctioned exactly is explained by another site. Which one? Fill in:

That belongs to

5. Reflection (2 points)

“If you don't pay, you're not the customer but the product.” Do you think that's true? Explain in 1–2 sentences:

Total points: ___ / 12

🔑 Answer key — Worksheet 2

1. The service sells ad space. The longer and more attentively you stay, the more ads it can show — your attention is the goods it earns money with (the attention economy).

2. Ad 1 (yes — “Sponsored”), Post 2 (no — a private post), Ad 3 (yes — “Ad”). Giveaways: labels like “Sponsored”, “Ad”, “Advertisement”, a “Learn more”/“Shop” button.

3. More time = more ads shown = more revenue. That is why “stay long” is the real goal of the ranking.

4. That belongs to Datenschutz verstehen / Understanding data protection (ad profile, targeting, real-time auction). This site stays on the incentive/business model.

5. Accepted: any reasoned position. A nuanced answer is that you are not a paying customer but you are not “sold” either — your attention is rented out for advertising.

Topic 3: Filter bubble & echo chamber Name: ______________________ Class: ______________________ Date: ______________________

Worksheet 3 — Sorted and amplified

1. Match the terms (3 points)

Connect each term to its explanation (draw lines):

a) Filter bubble→ a group where everyone thinks the same and confirms one another
b) Echo chamber→ when the feed lifts outrage to the top because it brings many reactions
c) Outrage amplification→ when the algorithm mostly shows you what matches your previous reactions

2. Explain the loop (3 points)

A filter bubble forms in a loop. Put the steps in the correct order (1–4):

  • ◯ You react more strongly to one kind of post.
  • ◯ The algorithm shows you even more of it.
  • ◯ The algorithm remembers this preference.
  • ◯ You see something different less often — the bubble narrows.

3. Why does outrage sell? (3 points)

Angry or outrage-inducing posts often get many reactions. Explain why they therefore land higher in the feed — and why that says nothing about whether they are true:

4. True or false? (3 points)

Tick the box and briefly correct the false statements:

  • ☐ true ☐ false — “A filter bubble only forms if you build it on purpose.”
  • ☐ true ☐ false — “If a post has many reactions, it is especially true.”
  • ☐ true ☐ false — “Tapping ‘not interested’ can make the feed more diverse again.”

5. Burst the bubble (2 points)

Name two things you can do so that your feed becomes more diverse again:

Total points: ___ / 14

🔑 Answer key — Worksheet 3

1. a→“mostly shows you what matches your previous reactions” (filter bubble = what you are shown/withheld) · b→“a group where everyone thinks the same and confirms one another” (echo chamber = social reinforcement) · c→“the feed lifts outrage to the top because it brings many reactions”.

2. Order: You react more strongly (1) → the algorithm remembers the preference (2) → shows more of it (3) → you see something different less often, the bubble narrows (4).

3. Many reactions (comments, shares, long dwell time) are a strong engagement signal; the algorithm rewards engagement, so the post is shown higher. Reach comes from reaction, not from correctness — the ranking does not check whether something is true. (Cross-link: Fake News verstehen covers the truth check.)

4. false (it often forms by itself through the loop) · false (many reactions ≠ true; only a lot of engagement) · true.

5. At least 2 of: tap “not interested” · deliberately follow diverse sources · switch to chronological · seek out opposing views · take breaks. Any sensible empowering strategy counts.

Topic 4: Outrage & my feed contract Name: ______________________ Class: ______________________ Date: ______________________

Worksheet 4 — Make the feed serve you again

1. Name the mechanism (3 points)

Describe in your own words why with some apps you “just keep scrolling” without noticing. What design trick is behind it?

2. Match the tools (3 points)

Connect each tool to what it does:

a) Switch to “chronological”→ you lose less time to constant interruptions
b) Reduce notifications→ the feed pulls you back less often and narrows more slowly
c) Follow diversely / “not interested”→ you see posts in time order instead of by engagement

3. Frame comparison pressure (3 points)

In feeds you often see only other people's best moments. Explain mechanically (not judgementally) why that gives a distorted picture:

4. My feed contract (3 points)

Write down three of your own rules for how you want to use your feed. Begin each with “I …”:

5. Reflection (2 points)

Which one rule is hardest for you — and why? Answer honestly in 1–2 sentences:

Total points: ___ / 14

🔑 Answer key — Worksheet 4

1. Accepted: the feed has no end and always loads new posts (“endless scrolling”); each new post promises a small reward, and there is no natural stopping point. Mechanics, not “weak willpower”.

2. a→“posts in time order instead of by engagement” · b→“pulls you back less often and narrows more slowly” · c→“you lose less time to constant interruptions”. (A sensible match is enough.)

3. What gets shared is usually highlights (selection bias); the algorithm additionally lifts especially appealing posts to the top. This creates the impression that everyone else's life is always perfect — although you only see a filtered selection. Empowering: this is a design effect, not a real comparison.

4. Accepted: three actionable, self-determined rules (e.g. “I switch to chronological”, “I put my phone in another room while studying”, “I also follow sources that disagree with me”).

5. Accepted: any honest, reasoned answer. This is about self-reflection, not a “correct” solution.

Class test — final assessment

Duration: 45 min · Points: 30 · Grade: per rubric below

The test does not fit into a 90-minute double lesson — recommended as its own lesson at the end of the weekly module (3 × 50 min) or as the close of the 5-week project.

Understanding social media — final test Name: ______________________ Class: ______________________ Date: ______________________ Points: ___ / 30

Part A: Multiple choice 1 pt each · 6 pts

1. What is the real goal of a feed algorithm?

  • To show the truest posts at the very top
  • To show the posts you are likely to stay with for a long time
  • To sort everything strictly by time
  • To show as few ads as possible

2. Why are many social networks free?

  • Because they live on donations
  • Because the state pays for them
  • Because they sell advertising using your attention
  • Because they need no money

3. What describes a filter bubble best?

  • A group where everyone shares the same opinion
  • The feed mostly shows you what matches your previous reactions
  • A technical fault in the app
  • A paid advertisement

4. Which of these signals is strongest for a ranking?

  • A post is shared
  • A post is viewed briefly
  • A post is liked
  • A post is reported and hidden

5. Why do outrage-inducing posts often land high in the feed?

  • Because they are always true
  • Because they trigger many reactions (a lot of engagement)
  • Because the app picks them to inform you
  • Because outrage is forbidden

6. What can make the feed more diverse again?

  • Liking even more of one kind
  • Using “not interested” and following diversely
  • Opening the app more often
  • Turning on all notifications

Part B: Short answer 2 pts each · 8 pts

7. Explain in 1–2 sentences what it means that a feed is “built and not neutral”:

8. Order the four engagement signals by their strength (strongest first):

9. What is the difference between a filter bubble and an echo chamber?

10. What does “the attention economy” mean — and why are services therefore often free?

Part C: Application 10 pts

11. Someone likes five animal videos in a row. Describe step by step how the feed changes afterwards — and name the technical term for this loop (3 pts):

12. Describe step by step how a single like can turn into a filter bubble — from the first reaction to the narrowed feed (4 pts):

13. Explain in your own words why “many reactions” is not the same as “true” (3 pts):

Part D: Reflection 6 pts

14. Discuss in 5–8 sentences: “Should apps be required to offer a chronological feed by default?” Give one argument for and one against, then take a reasoned position of your own.

🔑 Answer key for teachers — Class test

Part A: 1b · 2c · 3b · 4a · 5b · 6b

Part B:

7. “Built, not neutral” means: a piece of software decides, by fixed goals (engagement/dwell time), in what order you see posts — it is not a random or purely time-based list. (2 pts: a clear statement = 2 pts; partially correct = 1 pt)

8. Share → Comment → Like → Dwell time. Full marks for the correct order.

9. Filter bubble = what the algorithm shows or withholds (technical personalisation). Echo chamber = a social environment where everyone reinforces the same opinion. Both increase one-sidedness, but in different ways.

10. The attention economy = attention is the scarce goods competed for. Services are free because they turn your attention into money via advertising — the longer you stay, the more ads.

Part C:

11. Ideal answer: likes raise the weight of the “animals” category → the next batch contains more animal videos → you react again → other categories appear less. Technical term: personalisation loop / filter bubble. Points: mechanics (2), technical term (1).

12. Ideal answer: reaction (like) → the algorithm remembers the preference → shows more of it → more reactions → other topics disappear → the feed narrows (the diversity index drops). Points: order of the loop (2), “more of it” named (1), narrowing/diversity named (1).

13. Reach comes from engagement (comments, shares, dwell time), not from correctness. The ranking rewards reaction, not truth — outrage in particular generates many reactions even when the content is false. (3 pts for: engagement ≠ truth + the outrage example + the statement that the ranking does not check.)

Part D:

14. Full marks: one clear pro argument (e.g. more self-determination, less narrowing, transparency), one con argument (e.g. fewer relevant posts, harder to find, interference with the product) and a reasoned position of one's own. The depth of the argument is graded, not the position.

Grading rubric

PointsGrade (DE/AT)Grade (CH)Assessment
27 – 301 / Very good5.5 – 6.0Complete understanding, independent reflection, precise terminology.
23 – 262 / Good4.5 – 5.0Confident knowledge, minor gaps, reflection present.
18 – 223 / Satisfactory3.5 – 4.0Basics understood, reflection superficial.
14 – 174 / Sufficient3.0 – 3.4Key terms present, many gaps in application.
0 – 135 / Insufficient< 3.0Basic terms not understood — extra support recommended.

Grades follow the DE/AT 1–5 and CH 1–6 systems; map the point bands to your own grading scale as needed.

Weighting recommendation

Homework collection — 3 difficulty tiers

3 tasks per topic: Easy Medium Challenging. Answer hints are in expandable details directly underneath (collapsible on screen, always open when printed).

Every task is solvable without an account — no one has to create a profile or show their private feed. Where an app would help, there is an equivalent observation or research variant.

Topic 1 — Be the algorithm

Task HW 1.1 Easy

Explain to a family member in 3 sentences what a feed algorithm does. Then write down what question they asked.

🔑 Answer hint

Core: the algorithm sorts posts by how likely you are to react / stay long — not by truth or time. Accepted: own words + a documented follow-up question.

Task HW 1.2 Medium

Open the filter-bubble simulator on the site. Play two rounds: once feeding only one category, once reacting diversely. Note in 4–6 sentences how the diversity index differs. (No account, no login.)

🔑 Answer hint

Insight: one-sided reacting lowers the index and lets one category dominate; diverse reacting + “not interested” keeps it high. Accepted: a documented observation with numbers.

Task HW 1.3 Challenging

Research the steps candidate generation → scoring → re-ranking. Explain in 5–6 sentences what happens in each step.

🔑 Answer hint

Candidate generation: a pre-selection from millions of posts. Scoring: give each candidate a score for expected engagement. Re-ranking: fine-tune the order (e.g. diversity, fresh content). Bonus: a note that dwell time feeds in here.

Topic 2 — Attention is the currency

Task HW 2.1 Easy

Look at 5 posts on any public page and work out which of them are ads. Note the giveaway (e.g. “Sponsored”). (No account needed.)

🔑 Answer hint

Giveaways: labels “Sponsored/Ad/Advertisement”, a “Learn more” or “Shop” button, a company name as the sender. Accepted: 5 posts correctly classified.

Task HW 2.2 Medium

Explain in 4–6 sentences the “free app, paid advertising” business model. Why is “stay long” the goal?

🔑 Answer hint

Core: ad space is sold; more time = more ads shown = more revenue. Attention is the goods. Bonus: a note that the data mechanics behind it belong to data protection.

Task HW 2.3 Challenging

Research the term “attention economy”. Write a short paragraph (6–8 sentences): why is attention a scarce good, and who competes for it?

🔑 Answer hint

Attention is limited (each day has only 24 hours), content is abundant. Platforms, apps and advertising compete for this scarce resource. Accepted: a correct explanation in own words with an example.

Topic 3 — Filter bubble & echo chamber

Task HW 3.1 Easy

Explain in 3 sentences the difference between a filter bubble and an echo chamber.

🔑 Answer hint

Filter bubble = technical personalisation (what the algorithm shows/withholds). Echo chamber = a social group reinforcing the same opinion. Both narrow your view.

Task HW 3.2 Medium

Draw the filter-bubble loop as a comic in 3–4 panels (react → remember → more of it → narrowing).

🔑 Answer hint

Panel 1: a person likes a topic → Panel 2: the algorithm remembers the preference → Panel 3: shows more of it → Panel 4: other topics disappear. Accepted: correct order, creatively done.

Task HW 3.3 Challenging

Compare two public topic searches for the same term (e.g. on two different platforms or search engines). Note in 6–8 sentences which differences you notice and what might cause them. (No account needed.)

🔑 Answer hint

Accepted: a documented comparison with examples. Insight: different sorting/selection depending on the platform's ranking — even without a login it shows that “neutral results” are rare.

Topic 4 — Outrage & amplification

Task HW 4.1 Easy

Explain in 3 sentences why outrage-inducing posts often get more reach.

🔑 Answer hint

Outrage generates many reactions (comments, shares, long dwell time); the algorithm rewards engagement and shows such posts higher.

Task HW 4.2 Medium

Watch a public feed for 10 minutes (without reacting) and count how many posts try to “rile you up”. Describe in 4–6 sentences what you notice. (No account needed.)

🔑 Answer hint

Accepted: a documented observation. Insight: emotionally sharpened posts are frequent because they generate reactions. Important: no judging of individual people, only observing the mechanism.

Task HW 4.3 Challenging

Discussion text (max. 1 page): “Why does a lot of engagement say nothing about truth?” Bring in fake news and explain where detection must be separated from amplification.

🔑 Answer hint

Core: reach comes from reaction, not from correctness; the ranking checks no facts. Amplification (social media) and detecting false news (Fake News verstehen) are two different jobs. A reasoned position of one's own is required.

Topic 5 — Healthy, self-determined use

Task HW 5.1 Easy

Name 3 concrete tools that make your feed more self-determined, and write one sentence each on what they do.

🔑 Answer hint

E.g. switch to chronological (order by time), “not interested” (weaken a category), reduce notifications (pulled back less often), follow diversely (more perspectives).

Task HW 5.2 Medium

Write your own “feed contract” with 5 rules. Try it for a week and then note which rule worked best.

🔑 Answer hint

Accepted: 5 actionable, self-determined rules + an honest review after a week. Grade in an empowering way, not morally.

Task HW 5.3 Challenging

Discussion essay (max. 1 page): “Should schools make media literacy a compulsory subject?” Argue for and against.

🔑 Answer hint

Accepted: a balanced argument. Pro: participation, self-protection, informed media use. Con: a full timetable, the fast change of platforms, responsibility. A reasoned position of one's own is required.

Parent-letter template

You can adapt this template to your school and class. Replace the [placeholders in magenta] with your own details and print the template for your class. The letter deliberately contains no bans, only conversation starters.

[Your school]
[Address]
[Date]

To the parents of class [Year X]

Subject: Teaching unit “How do feeds & algorithms work?”

Dear parents,

over the coming [weeks / double lesson] your child's class will be looking at the question of who actually decides what appears in a social-media feed. Your child probably uses social media every day — we want them to understand, too, how a feed is sorted and how to make it serve them again. Whoever sees through the mechanics uses social media more confidently and more calmly.

Something that matters to us: we approach the topic factually and without wagging a finger. It is about mechanics, not bans or blame.

What your child will learn:

  • Why a feed is “built and not neutral” — and what attention has to do with it.
  • What an algorithm sorts and which signals (sharing, comment, like, dwell time) it rewards.
  • Why many services are free (the attention economy).
  • How filter bubbles and echo chambers form — and how to open them up again.
  • Why outrage is amplified, even though it says nothing about truth.
  • Which tools support healthy, self-determined use.

Material and source: We use the freely available learning platform social-media-verstehen.webhoch.com, provided by the Austrian agency Webagentur Hochmeir e.U. under a free licence (CC BY 4.0). The content is age-appropriate for [Years 6–9] and includes an interactive “filter-bubble simulator” that runs entirely without an account in the browser.

Three conversation starters for home (instead of bans):

  • Look at the filter-bubble simulator together — your child can explain to you how a feed narrows.
  • Ask openly: “What do you often see in your feed right now — and what might be causing that?” Without judging.
  • Talk about sleep and screen time as a shared agreement (e.g. no phone in the bedroom). The topic of “which data is collected along the way” is explored in depth by Datenschutz verstehen (data protection).

Important notes:

  • No social-media account is needed in the lesson; no private profiles or real names are shown. The site works without any sign-in.
  • It is used [on the projector / in the computer room / with school tablets].
  • If you have questions or concerns: [Your email address]

We are glad that your child is gaining this important basic skill — and we appreciate your support along the way.

Kind regards,
[Your name]
[Role / class teacher]

Curriculum mapping

This unit covers core learning areas of media literacy / digital literacy. In general, it maps to lower-secondary media-literacy and digital-literacy curricula (≈ Years 6–9, ages 11–15). The mapping below is adaptable to the German, Austrian and Swiss education standards for lower secondary. In Austria it fits the compulsory subject Digitale Grundbildung.

Curriculum references (general lower-secondary media/digital literacy; DE/AT “Digitale Grundbildung”, CH Lehrplan 21 “Medien und Informatik”)

Subject / areaCompetence / standardWorksheet
Digital literacy / media literacyExplain how recommendation and ranking systems work; recognise algorithmic personalisation1, 3
Media literacySpot advertising in the feed; situate reach critically; check content and sources2, 3
Language / EnglishExplain facts, argue, justify your own position1–4 (reflection parts)
Ethics / civicsResponsibility, opinion-forming and rules in the digital space (business models, outrage)2, 4
Computing (extension)Ranking as a prediction problem; signals, scoring, re-ranking1 (with HW 1.3)

Competence matrix (cognitive demand levels)

LevelNameExample from this course
Level IReproductionName the engagement signals; define filter bubble/echo chamber
Level IIApplicationDescribe the filter-bubble loop; spot ads
Level IIITransfer / evaluationA mandatory chronological feed? · justify “my feed contract”

Time allocation

VariantBreakdownRecommended for
Double lesson (90 min)Intensive workshop, all topics in excerpts, 2 worksheetsProject day, cover lesson, taster course
Weekly module (3 × 50 min)Day 1: algorithm & simulator; Day 2: attention & filter bubbles; Day 3: outrage, healthy use + testStandard lessons across one week
5-week project (5 × 90 min)One double lesson per topic, own research projects, final testIn-depth study, project work, gifted-and-talented support
Project week (5 × 4h)Days 1–4: deep dive into the topics with own research · Day 5: presentations + testThemed week, media camp

Where do I connect this?