Pathmaker

TemplatesAlgorithms and the Online Experience
beginner
Technical Systems
analytical

Algorithms and the Online Experience

Investigate how ranking, recommendation, advertising, and moderation mediate everyday online life.

6 weeks
5 hrs/week
30 total hours

About This Course

New Conditions: Fall 2026. A self-directed seminar and studio combining critical inquiry, practical experimentation, and critique.

Investigate how ranking, recommendation, advertising, and moderation mediate everyday online life. Final project: A platform field study and illustrated account of how visibility is shaped.

No specialist background required. Professional application: translate the final project into a case study explaining your research, decisions, execution, and limitations.

This is an original Pathmaker curriculum, not a university credential or affiliated school course.

Materials: spreadsheet or notebook, access to one platform through your own account, and a drawing or layout tool. No coding required.

Use invented posts for ranking exercises and anonymized observations for research. Do not collect private third-party data.

Final application: a research and visual-communication case study; this introductory course does not establish engineering competence.

Course structure: six weeks, five active study hours per week. Complete weekly readings, one documented artifact each week, and a final project.

External resources remain the work of their authors; links are supplied for study, and no institutional endorsement is implied. Required core readings are free to read; optional materials, software features, and physical production may incur costs.

Readings and links reviewed October 2, 2026. Historical university resources are labeled by their original year.

What You'll Learn

  • 1
    Explain and apply ranking.
  • 2
    Explain and apply feedback loop.
  • 3
    Explain and apply optimization objective.
  • 4
    Explain and apply moderation.
  • 5
    Produce a platform field study and illustrated account of how visibility is shaped.
  • 6
    Document sources, test assumptions, and revise work through critique.

Week by Week

1

Algorithms as rules

An algorithm is a procedure, not necessarily machine learning. A chronological feed sorts by time; a scored feed combines signals.

Work through a weighted example: recency 0–10 plus relevance 0–10 plus popularity 0–10, weighted 0.2, 0.5, and 0.3. Changing the weights changes whose posts become visible even when the content stays identical.

Worked example: two invented posts score (R=9,V=3,P=2) and (R=2,V=8,P=7). At weights (0.2,0.5,0.3), their totals are 3.9 and 6.5.

At (0.5,0.3,0.2), they are 5.8 and 4.8. A different objective changes the order without changing either post.

Use this example as a model for distinguishing a known calculation from an inference about a real platform.

Explore this week's curriculum

Assigned reading and resources

Google: Recommendation Systems

Required study: Read the overview and identify recommendation versus explicit search. Allow approximately 45 minutes for reading and notes, an estimate rather than the source's stated duration. Link reviewed October 2, 2026.

Free reading; optional tools or materials may cost extra · Approx. 45 minutes

Assessment and completion

Explain ranking in your own words, show where it appears in your artifact, and name one limitation. Check against the definition and required reading.

Use the activity's four-dimension rubric; revise any score below 2.

Key concepts

Ranking: Ordering items according to selected signals.

Signal: An observable input used to estimate relevance or another objective.

Studio activities

Week 1: Algorithms as rules

Create a table of ten invented posts with recency, relevance, and popularity scores from 0 to 10. Rank them using 0.2R + 0.5V + 0.3P; change the weights to 0.5R + 0.3V + 0.2P. Submit both rankings and a 250-word explanation of who gains visibility.

  1. Create a table of ten invented posts with recency, relevance, and popularity scores from 0 to 10.
  2. Rank them using 0.2R + 0.5V + 0.3P; change the weights to 0.5R + 0.3V + 0.2P.
  3. Submit both rankings and a 250-word explanation of who gains visibility.
  4. Record sources and permissions; label observations, assumptions, and invented examples.
  5. Review the rubric, obtain critique where possible, and document one substantive revision.
Success criteria

Submit the specified artifact and process record. Score each dimension 0–3: conceptual accuracy, evidence/method, execution, and reflection/revision. 0 = absent; 1 = partial or unsupported; 2 = clear and supported; 3 = precise, well tested, and critically explained.

Target at least 2 in every dimension before proceeding. This is self-assessment, not independently graded certification.

Reflection

  • How did ranking affect a specific decision in your study?
  • Which part of your result is supported by evidence, and which remains a hypothesis?
  • What alternative approach did you reject, and what tradeoff led to that choice?

Study guidance

  • Weekly schedule: 45 minutes required reading, 30 minutes concept notes, 2 hours 30 minutes studio/fieldwork, 45 minutes critique and revision, 30 minutes reflection and documentation. Total: 5 hours.
  • Study task: Read the overview and identify recommendation versus explicit search.
  • Use a paper or digital process notebook. Log date, question, source, method, observation, decision, and limitation.
  • If working alone, explain your result aloud and compare it to the rubric; an AI coach can support discussion but is not an independent assessor.
  • Required reading is a starting point: follow the specified section or navigation, not the entire linked course. Optional deeper research can extend the five-hour schedule.

Discuss with Professor Reed

Choose a question to discuss with Professor Reed. Copy it and paste it into your conversation with him. The copied text includes this week's lesson and assignment; add your own questions or explain where you are stuck.

Explain ranking and signal using my assignment. Ask me to explain the difference before giving feedback.

Use the worked example to test my reasoning. Change one condition and ask what I predict.

Review my process record against this week's criteria. Identify unsupported claims and suggest a manageable next test.

How did ranking affect a specific decision in your study?

Which part of your result is supported by evidence, and which remains a hypothesis?

What alternative approach did you reject, and what tradeoff led to that choice?

Ask me one question at a time to check my understanding, then explain any gaps.

Help me plan this assignment without doing the work for me.

5 hrs
2

Recommendation and feedback

Recommendation systems may use similarities between users, attributes of items, or both. An interaction is an ambiguous signal: watching longer might mean interest, confusion, or disagreement.

Repeated recommendations also change what someone can choose, making preference and exposure difficult to separate.

Worked example: two invented posts score (R=9,V=3,P=2) and (R=2,V=8,P=7). At weights (0.2,0.5,0.3), their totals are 3.9 and 6.5.

At (0.5,0.3,0.2), they are 5.8 and 4.8. A different objective changes the order without changing either post.

Use this example as a model for distinguishing a known calculation from an inference about a real platform.

Explore this week's curriculum

Assigned reading and resources

Google: Recommendation Systems

Required study: Use the course navigation to study content-based and collaborative filtering; draw one example of each. Allow approximately 45 minutes for reading and notes, an estimate rather than the source's stated duration. Link reviewed October 2, 2026.

Free reading; optional tools or materials may cost extra · Approx. 45 minutes

Assessment and completion

Explain feedback loop in your own words, show where it appears in your artifact, and name one limitation. Check against the definition and required reading.

Use the activity's four-dimension rubric; revise any score below 2.

Key concepts

Feedback loop: Outputs change behavior that becomes input for later recommendations.

Exposure bias: Observed preferences reflect which options a person was shown.

Studio activities

Week 2: Recommendation and feedback

Log the first ten recommended items in three sessions on your own account. Record time, item category, visible sponsorship, and your interaction. Submit 30 observations and a diagram of a possible feedback loop; label it as a hypothesis.

  1. Log the first ten recommended items in three sessions on your own account.
  2. Record time, item category, visible sponsorship, and your interaction.
  3. Submit 30 observations and a diagram of a possible feedback loop; label it as a hypothesis.
  4. Record sources and permissions; label observations, assumptions, and invented examples.
  5. Review the rubric, obtain critique where possible, and document one substantive revision.
Success criteria

Submit the specified artifact and process record. Score each dimension 0–3: conceptual accuracy, evidence/method, execution, and reflection/revision. 0 = absent; 1 = partial or unsupported; 2 = clear and supported; 3 = precise, well tested, and critically explained.

Target at least 2 in every dimension before proceeding. This is self-assessment, not independently graded certification.

Reflection

  • How did feedback loop affect a specific decision in your study?
  • Which part of your result is supported by evidence, and which remains a hypothesis?
  • What alternative approach did you reject, and what tradeoff led to that choice?

Study guidance

  • Weekly schedule: 45 minutes required reading, 30 minutes concept notes, 2 hours 30 minutes studio/fieldwork, 45 minutes critique and revision, 30 minutes reflection and documentation. Total: 5 hours.
  • Study task: Use the course navigation to study content-based and collaborative filtering; draw one example of each.
  • Use a paper or digital process notebook. Log date, question, source, method, observation, decision, and limitation.
  • If working alone, explain your result aloud and compare it to the rubric; an AI coach can support discussion but is not an independent assessor.
  • Required reading is a starting point: follow the specified section or navigation, not the entire linked course. Optional deeper research can extend the five-hour schedule.

Discuss with Professor Reed

Choose a question to discuss with Professor Reed. Copy it and paste it into your conversation with him. The copied text includes this week's lesson and assignment; add your own questions or explain where you are stuck.

Explain feedback loop and exposure bias using my assignment. Ask me to explain the difference before giving feedback.

Use the worked example to test my reasoning. Change one condition and ask what I predict.

Review my process record against this week's criteria. Identify unsupported claims and suggest a manageable next test.

How did feedback loop affect a specific decision in your study?

Which part of your result is supported by evidence, and which remains a hypothesis?

What alternative approach did you reject, and what tradeoff led to that choice?

Ask me one question at a time to check my understanding, then explain any gaps.

Help me plan this assignment without doing the work for me.

5 hrs
3

Advertising and incentives

Platforms select measurable proxies such as clicks, watch time, or purchases. These proxies do not automatically represent satisfaction or public value.

A sponsored placement, organic ranking, and a creator's distribution strategy are different mechanisms; label them separately in your map.

Worked example: two invented posts score (R=9,V=3,P=2) and (R=2,V=8,P=7). At weights (0.2,0.5,0.3), their totals are 3.9 and 6.5.

At (0.5,0.3,0.2), they are 5.8 and 4.8. A different objective changes the order without changing either post.

Use this example as a model for distinguishing a known calculation from an inference about a real platform.

Explore this week's curriculum

Assigned reading and resources

FTC: Disclosures 101

Required study: Read the disclosure guidance; separate sponsored content labels from ranking explanations. Allow approximately 45 minutes for reading and notes, an estimate rather than the source's stated duration. Link reviewed October 2, 2026.

Free reading; optional tools or materials may cost extra · Approx. 45 minutes

Assessment and completion

Explain optimization objective in your own words, show where it appears in your artifact, and name one limitation. Check against the definition and required reading.

Use the activity's four-dimension rubric; revise any score below 2.

Key concepts

Optimization objective: The measurable outcome a system is designed to improve.

Proxy metric: A measurable substitute for a goal that may not fully represent it.

Studio activities

Week 3: Advertising and incentives

Draw a four-party incentive map for users, creators, advertisers, and the platform. Annotate a sponsored post and an organic post. Submit a 400-word explanation separating disclosure requirements from speculative ranking behavior.

  1. Draw a four-party incentive map for users, creators, advertisers, and the platform.
  2. Annotate a sponsored post and an organic post.
  3. Submit a 400-word explanation separating disclosure requirements from speculative ranking behavior.
  4. Record sources and permissions; label observations, assumptions, and invented examples.
  5. Review the rubric, obtain critique where possible, and document one substantive revision.
Success criteria

Submit the specified artifact and process record. Score each dimension 0–3: conceptual accuracy, evidence/method, execution, and reflection/revision. 0 = absent; 1 = partial or unsupported; 2 = clear and supported; 3 = precise, well tested, and critically explained.

Target at least 2 in every dimension before proceeding. This is self-assessment, not independently graded certification.

Reflection

  • How did optimization objective affect a specific decision in your study?
  • Which part of your result is supported by evidence, and which remains a hypothesis?
  • What alternative approach did you reject, and what tradeoff led to that choice?

Study guidance

  • Weekly schedule: 45 minutes required reading, 30 minutes concept notes, 2 hours 30 minutes studio/fieldwork, 45 minutes critique and revision, 30 minutes reflection and documentation. Total: 5 hours.
  • Study task: Read the disclosure guidance; separate sponsored content labels from ranking explanations.
  • Use a paper or digital process notebook. Log date, question, source, method, observation, decision, and limitation.
  • If working alone, explain your result aloud and compare it to the rubric; an AI coach can support discussion but is not an independent assessor.
  • Required reading is a starting point: follow the specified section or navigation, not the entire linked course. Optional deeper research can extend the five-hour schedule.

Discuss with Professor Reed

Choose a question to discuss with Professor Reed. Copy it and paste it into your conversation with him. The copied text includes this week's lesson and assignment; add your own questions or explain where you are stuck.

Explain optimization objective and proxy metric using my assignment. Ask me to explain the difference before giving feedback.

Use the worked example to test my reasoning. Change one condition and ask what I predict.

Review my process record against this week's criteria. Identify unsupported claims and suggest a manageable next test.

How did optimization objective affect a specific decision in your study?

Which part of your result is supported by evidence, and which remains a hypothesis?

What alternative approach did you reject, and what tradeoff led to that choice?

Ask me one question at a time to check my understanding, then explain any gaps.

Help me plan this assignment without doing the work for me.

5 hrs
4

Moderation and visibility

Moderation can include removal, labeling, reduced distribution, and restrictions on accounts. Policy language describes intended rules, while enforcement evidence describes particular outcomes.

An appeal pathway should show notice, reason, response opportunity, review, and remedy.

Worked example: two invented posts score (R=9,V=3,P=2) and (R=2,V=8,P=7). At weights (0.2,0.5,0.3), their totals are 3.9 and 6.5.

At (0.5,0.3,0.2), they are 5.8 and 4.8. A different objective changes the order without changing either post.

Use this example as a model for distinguishing a known calculation from an inference about a real platform.

Explore this week's curriculum

Assigned reading and resources

Santa Clara Principles: Content Moderation

Required study: Study notice, transparency, and appeal principles and compare them with a platform policy. Allow approximately 45 minutes for reading and notes, an estimate rather than the source's stated duration. Link reviewed October 2, 2026.

Free reading; optional tools or materials may cost extra · Approx. 45 minutes

Assessment and completion

Explain moderation in your own words, show where it appears in your artifact, and name one limitation. Check against the definition and required reading.

Use the activity's four-dimension rubric; revise any score below 2.

Key concepts

Moderation: Rules and processes determining which content remains available or gains distribution.

Procedural accountability: The ability to understand, challenge, and review a decision process.

Studio activities

Week 4: Moderation and visibility

Compare the Santa Clara Principles with one platform's current public moderation policy. Create a notice-to-appeal journey and list three gaps or unanswered questions, citing the policy URL and retrieval date.

  1. Compare the Santa Clara Principles with one platform's current public moderation policy.
  2. Create a notice-to-appeal journey and list three gaps or unanswered questions, citing the policy URL and retrieval date.
  3. Record sources and permissions; label observations, assumptions, and invented examples.
  4. Review the rubric, obtain critique where possible, and document one substantive revision.
Success criteria

Submit the specified artifact and process record. Score each dimension 0–3: conceptual accuracy, evidence/method, execution, and reflection/revision. 0 = absent; 1 = partial or unsupported; 2 = clear and supported; 3 = precise, well tested, and critically explained.

Target at least 2 in every dimension before proceeding. This is self-assessment, not independently graded certification.

Reflection

  • How did moderation affect a specific decision in your study?
  • Which part of your result is supported by evidence, and which remains a hypothesis?
  • What alternative approach did you reject, and what tradeoff led to that choice?

Study guidance

  • Weekly schedule: 45 minutes required reading, 30 minutes concept notes, 2 hours 30 minutes studio/fieldwork, 45 minutes critique and revision, 30 minutes reflection and documentation. Total: 5 hours.
  • Study task: Study notice, transparency, and appeal principles and compare them with a platform policy.
  • Use a paper or digital process notebook. Log date, question, source, method, observation, decision, and limitation.
  • If working alone, explain your result aloud and compare it to the rubric; an AI coach can support discussion but is not an independent assessor.
  • Required reading is a starting point: follow the specified section or navigation, not the entire linked course. Optional deeper research can extend the five-hour schedule.

Discuss with Professor Reed

Choose a question to discuss with Professor Reed. Copy it and paste it into your conversation with him. The copied text includes this week's lesson and assignment; add your own questions or explain where you are stuck.

Explain moderation and procedural accountability using my assignment. Ask me to explain the difference before giving feedback.

Use the worked example to test my reasoning. Change one condition and ask what I predict.

Review my process record against this week's criteria. Identify unsupported claims and suggest a manageable next test.

How did moderation affect a specific decision in your study?

Which part of your result is supported by evidence, and which remains a hypothesis?

What alternative approach did you reject, and what tradeoff led to that choice?

Ask me one question at a time to check my understanding, then explain any gaps.

Help me plan this assignment without doing the work for me.

5 hrs
5

Investigating a platform

A feed diary is observational evidence rather than proof of an internal model. Account history, location, time, inventory, and platform experiments may explain differences.

Define one variable to change, a fixed observation window, and what would count as evidence against your explanation.

Worked example: two invented posts score (R=9,V=3,P=2) and (R=2,V=8,P=7). At weights (0.2,0.5,0.3), their totals are 3.9 and 6.5.

At (0.5,0.3,0.2), they are 5.8 and 4.8. A different objective changes the order without changing either post.

Use this example as a model for distinguishing a known calculation from an inference about a real platform.

Explore this week's curriculum

Assigned reading and resources

Google: Recommendation Systems

Required study: Review the retrieval/scoring/re-ranking sections; distinguish a toy model from observed platform behavior. Allow approximately 45 minutes for reading and notes, an estimate rather than the source's stated duration. Link reviewed October 2, 2026.

Free reading; optional tools or materials may cost extra · Approx. 45 minutes

Assessment and completion

Explain confounding variable in your own words, show where it appears in your artifact, and name one limitation. Check against the definition and required reading.

Use the activity's four-dimension rubric; revise any score below 2.

Key concepts

Confounding variable: An additional influence that makes causal interpretation difficult.

Causal claim: A statement that changing one factor produces a change in another.

Studio activities

Week 5: Investigating a platform

Pre-register a low-stakes observation: question, one changed behavior, fixed observation windows, and competing explanations. Run three sessions before and three after the change. Submit your log and explain why this cannot establish causation.

  1. Pre-register a low-stakes observation: question, one changed behavior, fixed observation windows, and competing explanations.
  2. Run three sessions before and three after the change.
  3. Submit your log and explain why this cannot establish causation.
  4. Record sources and permissions; label observations, assumptions, and invented examples.
  5. Review the rubric, obtain critique where possible, and document one substantive revision.
Success criteria

Submit the specified artifact and process record. Score each dimension 0–3: conceptual accuracy, evidence/method, execution, and reflection/revision. 0 = absent; 1 = partial or unsupported; 2 = clear and supported; 3 = precise, well tested, and critically explained.

Target at least 2 in every dimension before proceeding. This is self-assessment, not independently graded certification.

Reflection

  • How did confounding variable affect a specific decision in your study?
  • Which part of your result is supported by evidence, and which remains a hypothesis?
  • What alternative approach did you reject, and what tradeoff led to that choice?

Study guidance

  • Weekly schedule: 45 minutes required reading, 30 minutes concept notes, 2 hours 30 minutes studio/fieldwork, 45 minutes critique and revision, 30 minutes reflection and documentation. Total: 5 hours.
  • Study task: Review the retrieval/scoring/re-ranking sections; distinguish a toy model from observed platform behavior.
  • Use a paper or digital process notebook. Log date, question, source, method, observation, decision, and limitation.
  • If working alone, explain your result aloud and compare it to the rubric; an AI coach can support discussion but is not an independent assessor.
  • Required reading is a starting point: follow the specified section or navigation, not the entire linked course. Optional deeper research can extend the five-hour schedule.

Discuss with Professor Reed

Choose a question to discuss with Professor Reed. Copy it and paste it into your conversation with him. The copied text includes this week's lesson and assignment; add your own questions or explain where you are stuck.

Explain confounding variable and causal claim using my assignment. Ask me to explain the difference before giving feedback.

Use the worked example to test my reasoning. Change one condition and ask what I predict.

Review my process record against this week's criteria. Identify unsupported claims and suggest a manageable next test.

How did confounding variable affect a specific decision in your study?

Which part of your result is supported by evidence, and which remains a hypothesis?

What alternative approach did you reject, and what tradeoff led to that choice?

Ask me one question at a time to check my understanding, then explain any gaps.

Help me plan this assignment without doing the work for me.

5 hrs
6

Explaining mediated experience

Your report should distinguish documented platform statements, direct observations, and hypotheses. Illustrate one mechanism using a toy model and one actual observation.

Avoid claiming that your toy model reproduces a proprietary algorithm; its purpose is to make a possible mechanism intelligible.

Worked example: two invented posts score (R=9,V=3,P=2) and (R=2,V=8,P=7). At weights (0.2,0.5,0.3), their totals are 3.9 and 6.5.

At (0.5,0.3,0.2), they are 5.8 and 4.8. A different objective changes the order without changing either post.

Use this example as a model for distinguishing a known calculation from an inference about a real platform.

Explore this week's curriculum

Assigned reading and resources

Santa Clara Principles: Content Moderation

Required study: Revisit the principles and evaluate the accountability questions in your final report. Allow approximately 45 minutes for reading and notes, an estimate rather than the source's stated duration. Link reviewed October 2, 2026.

Free reading; optional tools or materials may cost extra · Approx. 45 minutes

Assessment and completion

Explain algorithmic literacy in your own words, show where it appears in your artifact, and name one limitation. Check against the definition and required reading.

Use the activity's four-dimension rubric; revise any score below 2. Completion requires all six weekly artifacts plus the final presentation, cited sources, and revision log.

Key concepts

Algorithmic literacy: The ability to reason about algorithmic systems and their limits.

Evidence classification: Separating direct observations, documented statements, and explanatory hypotheses.

Studio activities

Week 6: Explaining mediated experience

Deliver a 1,200-word field report or equivalent illustrated publication with a toy ranking model, feed evidence, incentive map, moderation analysis, five source notes, and an explicit limitations section.

  1. Deliver a 1,200-word field report or equivalent illustrated publication with a toy ranking model, feed evidence, incentive map, moderation analysis, five source notes, and an explicit limitations section.
  2. Record sources and permissions; label observations, assumptions, and invented examples.
  3. Review the rubric, obtain critique where possible, and document one substantive revision.
Success criteria

Submit the specified artifact and process record. Score each dimension 0–3: conceptual accuracy, evidence/method, execution, and reflection/revision. 0 = absent; 1 = partial or unsupported; 2 = clear and supported; 3 = precise, well tested, and critically explained.

Target at least 2 in every dimension before proceeding. This is self-assessment, not independently graded certification.

Final review: the artifact must match the brief, sources and permissions must be traceable, important limitations must be visible, and the case study must explain at least two revisions. A polished appearance cannot compensate for inaccurate claims or missing evidence.

Reflection

  • How did algorithmic literacy affect a specific decision in your study?
  • Which part of your result is supported by evidence, and which remains a hypothesis?
  • What alternative approach did you reject, and what tradeoff led to that choice?

Study guidance

  • Weekly schedule: 45 minutes required reading, 30 minutes concept notes, 2 hours 30 minutes studio/fieldwork, 45 minutes critique and revision, 30 minutes reflection and documentation. Total: 5 hours.
  • Study task: Revisit the principles and evaluate the accountability questions in your final report.
  • Use a paper or digital process notebook. Log date, question, source, method, observation, decision, and limitation.
  • If working alone, explain your result aloud and compare it to the rubric; an AI coach can support discussion but is not an independent assessor.
  • Required reading is a starting point: follow the specified section or navigation, not the entire linked course. Optional deeper research can extend the five-hour schedule.

Discuss with Professor Reed

Choose a question to discuss with Professor Reed. Copy it and paste it into your conversation with him. The copied text includes this week's lesson and assignment; add your own questions or explain where you are stuck.

Explain algorithmic literacy and evidence classification using my assignment. Ask me to explain the difference before giving feedback.

Use the worked example to test my reasoning. Change one condition and ask what I predict.

Review my process record against this week's criteria. Identify unsupported claims and suggest a manageable next test.

How did algorithmic literacy affect a specific decision in your study?

Which part of your result is supported by evidence, and which remains a hypothesis?

What alternative approach did you reject, and what tradeoff led to that choice?

Ask me one question at a time to check my understanding, then explain any gaps.

Help me plan this assignment without doing the work for me.

5 hrs

Topics Covered

fall 2026
new conditions
technical systems
algorithms and the online experience

More in This Track

Ready to begin?

Start your personalized curriculum today — free to try, no credit card needed.