**The most surprising finding is how close to deterministic the dataset is. Weekday daily usage almost fully tracks wellbeing, yet weekend extra hours have no relationship with anything.** **1. The correlations are far stronger than real survey data usually shows.** Across all 4,500 students, daily usage hours correlate at **-0.85 with the Mental Health Index**. Other correlations with usage are -0.72 with sleep hours, -0.71 with GPA, -0.68 with sleep quality and +0.75 with perceived stress. The outcomes also move together: sleep and mental health at +0.77, stress and mental health at -0.77, and GPA and mental health at +0.66 (steps 3 and 8). Behavioural survey data rarely gives -0.85 between a self-reported habit and a wellbeing score. This is the ordinary expectation (more usage, worse outcomes) pushed to an implausible degree. **2. The three impact groups barely overlap.** The averages by `Overall_Impact` show a clean gradient (step 1): | Overall_Impact | Students | Avg daily hours | Avg mental health | Avg GPA | Avg sleep (hrs) | |---|---|---|---|---|---| | Beneficial | 3,681 | 4.4 | 84.3 | 3.55 | 7.0 | | Neutral | 654 | 8.5 | 63.4 | 3.00 | 5.5 | | Negative | 165 | 12.3 | 47.1 | 2.54 | 4.2 | About 82% of students are labelled "Beneficial", and they are simply the lowest-usage group. Only 165 students (about 3.7%) are labelled "Negative". **3. Weekend extra hours are unrelated to everything.** - `Weekend_Extra_Hours` correlates **-0.01 with daily usage**. - It correlates 0.004 with mental health, 0.012 with GPA and 0.017 with sleep (steps 8 and 9). If weekend use came from the same underlying habit, it should track weekday use. Instead it looks like an independent random variable. **4. Age is also irrelevant.** Age correlates only 0.03 with GPA (step 9), even though the card lists age as a key segmentation variable. **How to read this:** Tight links between a handful of variables, alongside variables with essentially zero signal, suggest the data was generated from a formula driven mainly by daily usage. If so, the "dose-response" may be partly a construction of the data rather than a real-world effect. The data is also observational, so even a real relationship wouldn't show that usage causes the poorer outcomes. Treat the headline of "limit daily hours to protect sleep, GPA and mental health" with caution until it is checked against independent data.