Media & Information Literacy Grade 6 Algorithms and Personalisation

Grade 6 Media & Information Literacy Study Guide: Algorithms and Personalization

1. The Driving Question

"Why does YouTube keep recommending videos of my favorite soccer player, but my friend sees makeup tutorials instead? How does the app ‘know’ what I like—and can it ever get it wrong?" If the internet is supposed to show everyone the same information, why does your TikTok feed look totally different from your grandma’s? And if these apps are so smart, why do they sometimes suggest weird or even upsetting stuff?

2. The Core Idea — Built, Not Listed

Imagine you’re at a school cafeteria, and the lunch lady, Ms. Rivera, always remembers your order. If you pick pizza every Monday, she starts putting a slice on your tray before you even ask. But one day, you’re sick of pizza and want a salad—yet she keeps handing you pepperoni. That’s kind of how algorithms work: they predict what you’ll like based on what you’ve done before, but they don’t always notice when you change your mind.

Here’s how it happens online: 1. You leave clues—every like, search, or video you watch is a "data point" (like telling Ms. Rivera you love pizza).
2. The algorithm sorts you—it groups you with other people who like the same things (e.g., "kids who watch soccer highlights also watch sports documentaries").
3. It guesses what you’ll do next—just like Ms. Rivera assumes you’ll want pizza again, YouTube assumes you’ll want more soccer videos.
4. It tests its guesses—if you ignore a recommendation, the algorithm adjusts (like Ms. Rivera noticing you didn’t eat the pizza and trying a burger next time).

But here’s the catch: algorithms don’t understand you. They’re just math—patterns and probabilities. If you only watch funny cat videos, the algorithm won’t know you’re also interested in space documentaries unless you tell it (by searching or clicking). And if you accidentally watch a weird conspiracy video, it might start recommending more, even if you didn’t mean to.

Key Vocabulary:
- Algorithm: A set of step-by-step rules a computer follows to solve a problem or make a decision. Example: Netflix’s algorithm decides which shows to recommend by comparing your watch history to other users’—like a librarian who notices you checked out Harry Potter and suggests Percy Jackson next.
- Personalization: When an app or website changes what it shows you based on your past behavior. Example: Spotify’s "Discover Weekly" playlist is personalized—it won’t include country music if you only listen to hip-hop, even if country is popular that week.
- Data Point: A single piece of information about you or your behavior. Example: Every time you pause a YouTube video, that’s a data point the algorithm uses to guess if you’re bored or interested.
- Filter Bubble (Grade 6+): When personalization limits what you see online, trapping you in a "bubble" of content that matches your existing interests—so you might miss important news or different perspectives. Example: If you only follow animal rescue accounts on Instagram, you might never see posts about climate change or local elections.

3. Assessment Translation

How This Appears on State Tests (Grade 6):
- Multiple Choice: Questions about how algorithms work, with distractors that confuse personalization with randomness or human curation.
- Example: "Why does TikTok show you dance videos after you watch a few?"
- A) The app’s employees pick videos for you. (Distractor: confuses algorithms with human editors.)
- B) The app uses math to predict what you’ll like based on your past behavior. (Correct.)
- C) The videos are chosen randomly. (Distractor: ignores personalization.)
- D) The app only shows videos with the most likes. (Distractor: confuses popularity with personalization.) - Short Constructed Response: Explain how personalization can be helpful and limiting.
- Example: "Describe one way personalization makes your online experience better and one way it might cause problems. Use an example from a real app." - Proficient Response: "Personalization is helpful because it saves time—like when Spotify makes a playlist of songs I’ll probably like. But it can also be a problem because it might not show me new things, like if I only see soccer videos and miss out on learning about other sports." - Developing Response: "It’s good because it shows me stuff I like. It’s bad because it’s creepy." (Lacks examples and doesn’t explain the trade-off.)

What Teachers Look For:
- Evidence of understanding: Does the student explain how algorithms use data (e.g., "it notices I watch cooking videos and suggests more")? - Critical thinking: Does the student recognize both benefits (convenience) and drawbacks (filter bubbles)? - Specificity: Are examples named (e.g., "YouTube," "Amazon recommendations")?

Model Proficient Response (Short Answer):
"Algorithms personalize my Instagram feed by tracking which posts I like or save. For example, if I like a lot of baking videos, Instagram will show me more of those. This is helpful because I see content I enjoy, but it can also be limiting because I might miss posts about other topics, like science or news. It’s like if a friend only ever talked to me about baking—I’d learn a lot about cakes but not much else."

4. Mistake Taxonomy

Mistake 1: Confusing Algorithms with Human Editors
- Question: "How does YouTube decide which videos to recommend to you?" - Common Wrong Answer: "YouTube employees watch all the videos and pick the best ones for you." - Why It Loses Credit: The question asks about how recommendations work, not who makes them. This answer ignores the role of algorithms.
- Correct Approach: "YouTube’s algorithm looks at which videos you’ve watched, liked, or searched for in the past. It compares your behavior to other users’ and guesses what you’ll want to watch next. For example, if you watch a lot of Minecraft videos, it might recommend similar gaming content."

Mistake 2: Assuming Personalization is Always Accurate
- Question: "Your friend says, ‘If I watch one video about sharks, YouTube will only show me shark videos forever.’ Is this true? Explain." - Common Wrong Answer: "Yes, because algorithms are really smart and know exactly what you like." - Why It Loses Credit: This overestimates the algorithm’s accuracy. Algorithms predict but can be wrong or over-narrow.
- Correct Approach: "Not exactly. Algorithms start with guesses, but they adjust based on what you actually click on. If you watch one shark video but then ignore the next five, YouTube will stop recommending them. However, if you keep watching shark videos, it might assume you only like sharks and stop showing you other topics."

Mistake 3: Ignoring the "Filter Bubble" Problem
- Question: "Explain one way personalization might affect what you learn online." - Common Wrong Answer: "It’s good because you only see stuff you like." (Lacks critical thinking about drawbacks.) - Why It Loses Credit: The question asks for an effect, not just an opinion. A complete answer should acknowledge both pros and cons.
- Correct Approach: "Personalization can create a ‘filter bubble,’ where you only see content that matches your existing interests. For example, if you only watch funny animal videos, you might miss important news or different perspectives. This can make it harder to learn about new topics or understand other people’s views."

5. Connection Layer

1. Within Media Literacy → Misinformation: Algorithms can accidentally spread false information if they prioritize engagement (e.g., shocking headlines) over accuracy. Why it matters: Understanding how algorithms work helps you spot why a fake news story might go viral—it’s not because it’s true, but because it’s designed to get clicks.
2. Across Subjects → Math (Probability): Algorithms use probability to predict your behavior—like how a weather app calculates a 70% chance of rain. Why it matters: The math behind personalization is the same math that predicts outcomes in science or sports stats.
3. Outside School → Shopping Recommendations: Ever notice how Amazon suggests products "based on your browsing history"? That’s an algorithm at work—just like with videos, but for stuff to buy. Why it matters: Next time you see "Customers who bought this also bought…," you’ll know it’s not magic—it’s math tracking what people like you tend to purchase.

6. The Stretch Question

"If an algorithm only shows you content it thinks you’ll like, could it ever make you less curious? What would happen if you spent a week only clicking on things the algorithm recommends—would you discover anything new, or just get stuck in the same topics?"

Pointer Toward the Answer:
Algorithms are designed to keep you engaged, which often means showing you more of what you already like. But curiosity thrives on unexpected discoveries—like stumbling on a book in the library that’s nothing like what you usually read. If you only follow the algorithm’s suggestions, you might miss out on learning about history, art, or science just because you’ve never clicked on those topics before. Some apps (like Spotify’s "Discover Weekly") try to balance this by mixing familiar and new content, but it’s up to you to break out of the bubble sometimes. What’s one topic you’ve never explored online that you might search for this week?