The sample above has 15 rows spread across the Student_ID range (STU_2024398 to STU_2028135). A second view of 15 rows starts at STU_2024000. All 20 columns are in the dataset, but the sample shows only the 16 analytical ones. The four ingestion metadata columns aren't displayed. **What the rows show** - **Grain:** each row is one student with a unique `Student_ID`, covering demographics (age, gender, academic level), usage habits and outcomes. - **Demographics:** ages run from 15 to 23. Most students are High School or Undergraduate, with one Postgraduate. Genders are mostly Male and Female, plus one Non-Binary. - **Usage:** `Primary_Platform` includes YouTube, TikTok, Instagram, Reddit, Snapchat, LinkedIn and X (Twitter). Smartphones dominate, with a few Tablet and Laptop/PC users. `Daily_Usage_Hours` ranges from about 1.2 to 10.9 in the sample, and `Weekend_Extra_Hours` from 0.4 to 3.1. - **Late-night use:** `Late_Night_Usage` is a True/False flag. - **Comparison behavior:** `Social_Comparison_Frequency` takes values such as Never, Rarely, Sometimes, Frequently and Always. - **Outcomes:** these are sleep duration and quality, `Perceived_Stress_Score`, `Mental_Health_Index`, `Academic_Performance_GPA`, and an `Overall_Impact` label of Beneficial, Neutral or Negative. **Patterns visible in these rows (not statistically tested)** - Heavy daily users show worse outcomes. For example, STU_2025378 (10.4 h/day, TikTok) has 4.8 h of sleep, a sleep quality score of 1, stress of 29, a mental health index of 56 and a GPA of 2.37, and is labelled Negative. STU_2024002 (10.9 h/day) is labelled Neutral with 5.2 h of sleep. - Light users tend to do better. STU_2026053 (3.1 h/day) sleeps 8.9 h, has stress of 3, a mental health index of 95, a GPA of 4.0, and is labelled Beneficial. - Most rows are labelled Beneficial, so the label looks skewed toward positive. These are only 15 rows, so treat the patterns as illustrations. To test them across all 4,500 rows, I can compare outcomes by usage-hour bands, late-night use or platform.