{
    "action": "final",
    "content": "## What this dataset is about\n\nThis dataset, **impact-of-social-media-on-life**, is an observational survey-style dataset of **4,500 students**. It looks at how social media habits relate to wellbeing and academic outcomes. It has a single table, `raw.Social_media_impact_on_life`, with 20 columns and no joins to other tables.\n\nThe columns fall into a few groups:\n\n- **Student profile:** `Age`, `Gender`, `Academic_Level`.\n- **Social media usage:** `Primary_Platform`, `Daily_Usage_Hours` (0.9 to 14.0), `Weekend_Extra_Hours` (0.0 to 4.5), `Device_Type`, `Late_Night_Usage` (yes/no), and `Social_Comparison_Frequency`.\n- **Outcomes and wellbeing measures:** `Sleep_Duration_Hours`, `Sleep_Quality_Score`, `Perceived_Stress_Score`, `Mental_Health_Index`, `Academic_Performance_GPA`, and a categorical `Overall_Impact` label.\n- **Ingestion metadata:** `_ingestion_timestamp`, `_batch_id`, `_source_file`, `_source_system`. These are pipeline bookkeeping, not part of the survey.\n\nIts likely use is to inform digital-wellbeing decisions. Examples are which student segments to prioritize for outreach, whether to recommend limits on daily or late-night use to protect sleep and GPA, and which platforms or devices carry higher stress or mental-health risk. Because the data is observational, relationships in it show association, not cause.\n\n## What a single row represents\n\nEach row is **one student**, identified by a unique `Student_ID`. The dataset card lists this as the table's grain and key, with one row per `Student_ID`. A row holds that student's demographics, usage habits, and wellbeing and academic measures.\n\nThe two \"temporal\" columns, `Daily_Usage_Hours` and `Weekend_Extra_Hours`, are durations of usage. They are not dates, so the dataset has no time dimension for trend analysis. It is a cross-sectional snapshot.\n\nThese points come from the dataset card and its ontology notes. I did not run any queries against the data this turn."
}