Lesson 4: Modeling Amplification – How Platforms Shape Visibility

• How does engagement influence what appears in a user’s feed?
• What happens after someone likes, comments, or watches longer?
• Can small differences in engagement create large differences in visibility?
• Are algorithmic systems neutral, or do they reinforce certain behaviors?
• How does amplification reshape influence over time?

Algorithm
Feed
Amplification
Feedback loop
Signal
Ranking
Visibility
Reinforcement
Engagement signal
Recommendation
Distribution

HS ETS1-2
Analyze a complex real-world problem and model system behavior.

Science and Engineering Practices:
Developing and Using Models
Constructing Explanations
Analyzing and Interpreting Data

Crosscutting Concepts:
Systems and System Models
Cause and Effect
Stability and Change

• Interpreting system diagrams
• Analyzing multi-step cause-and-effect relationships
• Explaining feedback loops in written form
• Evaluating hypothetical scenarios using structured reasoning

Students practice translating conceptual systems into labeled diagrams and written explanations, mirroring analytical tasks found in standardized assessments.

Day 1 – From Metrics to Mechanism

Begin with a concrete scenario:

Two posts are published at the same time.
Post A receives slightly higher engagement in the first hour.

Students predict:

What happens next?

Students reason verbally first.

Class constructs a simple principle:

If engagement increases, visibility increases.
If visibility increases, engagement may increase further.

This becomes the foundation of a feedback loop.

Purpose:
Bridge proportional reasoning to system dynamics.

DOK: 2 – Identify causal relationships.

Day 2 – Diagramming the Feedback Loop

Students create a labeled model:

User Interaction ? Engagement Signal ? Algorithm Ranking ? Increased Visibility ? More Interaction

Students annotate:

Where does amplification occur?
Where could suppression occur?

They test the model against scenarios:

What if early engagement is weak?
What if engagement is concentrated in one subgroup?

Purpose:
Make algorithmic amplification visible as a structured system.

DOK: 3 – Develop and apply a systems model.

Day 3 – Scaling Effects and Stability

Students explore how small differences in engagement rate (from Segment 3) could produce exponential visibility differences over time.

Discussion includes:

Does amplification create fairness?
What kinds of content benefit most from feedback loops?
What kinds of content might be suppressed?

Students revise their diagram to include:

Initial conditions
Time
Reinforcement intensity

Purpose:
Introduce stability versus runaway amplification.

DOK: 3 – Analyze dynamic system behavior.

Optional Day 4 – Comparing Platforms

Students compare two hypothetical platform rules:

Platform A boosts any post with high engagement.
Platform B boosts posts only if engagement is sustained over time.

Students analyze how these different rules produce different content ecosystems.

Purpose:
Highlight how design decisions shape system outcomes.

DOK: 3

Students connect amplification to real experiences:

Why certain trends dominate feeds
Why extreme or emotional content spreads rapidly
Why smaller creators struggle to break through

Students begin to see how algorithm design shapes cultural conversation and exposure.

• Algorithms show content randomly.
• Algorithms are neutral tools.
• Good content naturally rises without reinforcement.
• Engagement and amplification are separate processes.
• Visibility is evenly distributed across users.

Teacher moves include pushing students to identify feedback loops explicitly rather than vaguely describing “the algorithm.”

• Provide partially completed feedback loop templates.
• Offer guided diagram labels for students who need structure.
• Allow verbal explanation before written modeling.
• Extension: Have students propose a modification to the loop and predict its effects.

Formative Assessments:

• Labeled feedback loop diagram
• Written explanation of amplification mechanism
• Scenario analysis response

Exit Ticket Prompt:

Explain how a small difference in engagement rate could lead to a large difference in visibility over time.

Evaluation Criteria:

Accuracy of system model
Clarity of cause-and-effect reasoning
Ability to apply proportional reasoning
Depth of explanation