← CS Unplugged
Algorithms Data Ethics 30–45 mins Groups of 4–6 No screens needed

How Does a Social Media Algorithm Decide What You See?

A CS Unplugged activity where students simulate a social media feed algorithm using cards — exploring engagement loops, filter bubbles, and data ethics without a screen in sight.

The UK government has announced a ban on social media for under-16s. But before discussing whether that is right or wrong — do your students actually know how social media decides what appears on their screen?

It is not random. It is not chronological. It is an algorithm — and understanding how it works is one of the most important things a young person can learn about the digital world.

This activity lets students simulate a social media feed algorithm using cards and coloured dots. No accounts, no screens, no phones — just a surprisingly illuminating set of decisions.

📋 What You Need

20 index cards, each representing a piece of content — write a brief description on each: a funny video, a news story, an advert, a post from a friend, a political opinion, a celebrity photo

Sticky labels or coloured dots to represent user preferences

Groups of 4–6 students (works equally well as a whole-class activity)

Approximately 30–45 minutes

🎮 The Activity

Round 1

Chronological Feed 5 minutes

Shuffle the cards and deal them out in order. Everyone sees the same content in the same sequence. This is how social media used to work in the early days.

Discuss

  • • Is this fair? Is it useful?
  • • What are the problems with showing everyone the same thing in the same order?
Round 2

Introducing Weights 10 minutes

Assign each student a profile: one loves sport, one loves music, one is interested in politics, one just wants to see friends' posts. Give each card a score from 1–3 for how relevant it is to each profile. Each student sorts their cards by score, highest first. This is their personalised feed.

Discuss

  • • How did the feeds differ between students?
  • • Who decided what scored highly? Who wrote those rules?
  • • Could the scoring system ever be unfair?
Round 3

The Engagement Trap 10 minutes

Remove the scoring system. Students now choose which cards they would stop and look at — whichever grab their attention. Mark those cards. Reshuffle and deal again, but this time cards that were previously marked are dealt first.

Discuss

  • • What kinds of content kept getting promoted?
  • • Was it the most accurate? The most useful? The most extreme?
  • • Who benefits from keeping you engaged?

💡 What Just Happened?

Students just simulated the core mechanism behind every major social media platform.

The algorithm does not show users what is true, or what is good for them. It shows what keeps them looking — because every second spent on the platform is a second it can show an advert.

Engagement-based algorithm

The more a user interacts with a type of content, the more of it they see.

Filter bubble

The feed narrows over time until it mostly confirms what you already think.

Feedback loop

Behaviour shapes the algorithm; the algorithm shapes behaviour.

💬 Big Discussion Questions

?

Who is responsible for what the algorithm promotes — the platform, the users, or both?

?

Could you design a fairer algorithm? What would you optimise for instead of engagement?

?

Should algorithms be transparent — should users be able to see why they are being shown something?

?

The UK government has announced a ban on social media for under-16s. Is understanding how algorithms work a better solution, a worse one, or something that should happen alongside the ban?

These work well as a written reflection, class debate, paired discussion, or homework task.

📚 Curriculum Links

Topic Connection
Algorithms How rules and conditions produce outputs from inputs
Data and its uses How platforms collect and act on behavioural data
Ethical and legal issues Who is responsible for algorithmic outputs
Artificial intelligence How machine learning refines recommendations over time
Social effects of computing Filter bubbles, misinformation, and platform design

🚀 Extension Activities

GCSE

Write pseudocode for a simple recommendation algorithm that takes user history as input and outputs a ranked feed.

A Level

Research and compare the Facebook News Feed, YouTube recommendations, and TikTok's For You page. What are the ethical implications of each?

All students

Spend one day noting every time an algorithm makes a decision for you — Netflix, Spotify, Google, maps. Discuss findings as a class.

Want more activities like this?

New CS Unplugged activities are added regularly — all written by a Senior Examiner and Head of CS, with full curriculum links.