Classwork 8

Databases - Social Media Analytics

Social Media Survey

🎯 Purpose of this Activity

Extract data from you (students)

By submitting the survey, you generate raw rows in a database table — just like how platforms quietly log every interaction you make.

• Behind the scenes, each of your clicks becomes a database entry — a row born from your activity.
• The survey is anonymous — we analyze patterns, not individuals.

Social Media Analytics

ETL Workflow

• Your role: You submit the survey — that generates raw data rows, just like a platform logging user behavior.

• My role: I act as the data system. As we walk through the results, I’ll point out where I use
  filter(), select(), and left_join() to Transform your responses and Load a clean analysis table we can analyze.

• Big idea: Your responses become a relational dataset, and then we join in algorithm rules — the same way TikTok, Instagram, or Reddit enrich user logs before ranking your feed.

Feed Algorithm

“Do you think your feed reflects your choices — or the platform’s choices about you?”

• We’ll look at platform behavior patterns using your data:
  minutes, posting activity, device type, and interaction style.
• Then we’ll run an algorithm simulation using a random card draw system to show how the same minutes can be valued very differently depending on the platform’s ranking logic.

🧠 Feed Algorithm Interpretation Using Card Mechanics

To illustrate how platforms differ in engagement efficiency, we treat each minute of screen time like drawing a card:

• Rank (2–10, J, Q, K, A) → how strong your engagement signal was
  
• Suit (platform personality) → how that platform interprets and amplifies that engagement

You’ll see that not all platforms reward the same behavior equally — some love viral spikes, some love steady engagement.

🎭 Platform Algorithm Personalities — Rank Bias Model

Each platform doesn’t just apply a fixed multiplier — it biases specific rank categories (Face, Ace, Mid, etc.).
This models how TikTok boosts viral spikes, Reddit rewards slower thread engagement, and YouTube locks users into deep watch sessions.

Platform	Algorithm Personality	Rank Preference Pattern
TikTok	♠ Strategic viral retention — aggressively hunts viral spikes	Face → Ace → 10 (very spike-oriented, punishes low/no engagement)
Instagram / Threads	♦ Aesthetic/status curation — polished trend amplification	10 → Face → Mid (reward polished content, suppress low-value scroll)
YouTube	♠ Strategic deep retention — optimizes long session depth	Ace → 10 → Mid (not chaotic, but strong on intentional consumption)
Reddit / Discord	♣ Community depth — favors threads, steady engagement over spikes	Mid → Ace → 10, downweights Face spikes (less about virality)
Snapchat	♥ Emotional burst algorithm — high volatility, short attention loops	Face spikes allowed, but low ranks common (chaotic scroll-emotion mix)
X/Twitter	♣ Fast discourse loop — rewards mid-level interaction & repost energy	Mid + Face modest bias, suppresses dead scrolling
Facebook	♦ Legacy ranking — mild engagement shaping, mostly neutral	Slight mid bias, nearly flat weighting otherwise
None	Neutral baseline — no algorithm shaping	All rank_adj = 1.00 (control group)

🎮 Rank Interpretation (Base Weights Before Platform Influence)

Rank Category	Base Weight	Meaning in Attention Model
Joker (0)	1.00	Passive drift — counted time, no engagement signal
2–4 (Low ranks)	1.00	Glance-scroll, content seen without response
5–9 (Mid ranks)	1.05	Light scrolling or minor interaction — retention signal
10	1.10	High interest moment — actively watching or reading
Ace (A)	1.25	Intentional engagement — searching, saving, deep dwell
Face Cards (J/Q/K)	1.50	Emotional/viral spike — strong reaction triggers

🚩 Raw minutes = counting cards without considering rank or suit.
🎯 Algorithm-adjusted attention = calculating your true score based on both rank (quality of engagement) and suit (type of platform algorithm).

In other words:

• A 20-minute session on Instagram Reels (♦ Ace) scores high adjusted attention.

• A 40-minute YouTube autoplay binge (♠ Joker) has high raw minutes but almost no algorithm-weighted value.

• Two users can spend the same time, but their card draws (attention efficiency) differ by platform.

  • TikTok may reward spike engagement, while Reddit rewards steady browsing.
  • YouTube may ignore your first 10 minutes, but heavily weight the moment you intentionally engage.

✅ Key Takeaway

Algorithms don’t just count your time — they interpret and score it.
Some platforms are time sinks, while others are attention amplifiers — even if the minutes look the same.

Questions

1. If platforms score and rank your behavior in secret, should users be allowed to know their score — or is it okay for the system to stay hidden?

2. If a platform is designed to keep you scrolling even when the content is low-value, is it ethical to design for addiction instead of meaningful use?

Discussion

Welcome to our Classwork 8 Discussion Board! 👋

This space is designed for you to engage with your classmates about the material covered in Classwork 8.

Whether you are looking to delve deeper into the content, share insights, or have questions about the content, this is the perfect place for you.

If you have any specific questions for Byeong-Hak (@bcdanl) or peer classmate (@GitHub-Username) regarding the Classwork 8 materials or need clarification on any points, don’t hesitate to ask here.

Let’s collaborate and learn from each other!