# Representative sample of the dataset The table is a stratified sample of 16 rows, 4 from each `sleep_debt_category`. The dataset has 8,500 rows in total (647 + 2,004 + 4,462 + 1,387). Each user appears once, so each row is one user. | user_id | age | gender | occupation | chronotype | bedtime app | bright. % | BL filter | caffeine after 5pm (mg) | activity (min) | latency (min) | sleep hrs | deep % | REM % | snoozes | fatigue | category | |---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---| | USR-04718 | 65 | Male | Remote Tech | Morning Lark | YouTube | 49 | 1 | 170 | 11 | 64.5 | 7.63 | 14.5 | 16.1 | 4 | 4.1 | Mild Deficit | | USR-05758 | 30 | Male | Freelance / Creative | Morning Lark | Instagram / Reddit | 50 | 0 | 0 | 24 | 41.3 | 6.90 | 20.8 | 21.3 | 3 | 2.9 | Mild Deficit | | USR-06746 | 64 | Female | Corporate 9-to-5 | Intermediate | YouTube | 46 | 1 | 0 | 0 | 38.4 | 7.06 | 24.0 | 23.7 | 2 | 1.7 | Mild Deficit | | USR-07693 | 26 | Male | Corporate 9-to-5 | Intermediate | Messaging / Chat | 63 | 1 | 71 | 49 | 29.1 | 7.37 | 24.4 | 14.6 | 2 | 2.7 | Mild Deficit | | USR-00217 | 27 | Male | Corporate 9-to-5 | Morning Lark | YouTube | 63 | 0 | 0 | 52 | 41.9 | 5.53 | 19.2 | 18.2 | 3 | 4.6 | Moderate Debt | | USR-01287 | 21 | Female | Corporate 9-to-5 | Intermediate | Instagram / Reddit | 37 | 1 | 104 | 45 | 49.3 | 5.00 | 18.6 | 21.1 | 4 | 6.7 | Moderate Debt | | USR-01361 | 53 | Male | Corporate 9-to-5 | Night Owl | News / Reading | 58 | 1 | 0 | 57 | 47.6 | 4.39 | 21.6 | 21.7 | 5 | 6.7 | Moderate Debt | | USR-08354 | 34 | Male | Healthcare / Shift Worker | Night Owl | Messaging / Chat | 10 | 1 | 28 | 36 | 24.3 | 3.97 | 21.3 | 21.3 | 3 | 5.7 | Moderate Debt | | USR-00858 | 31 | Female | Student | Morning Lark | YouTube | 54 | 1 | 40 | 22 | 36.7 | 8.29 | 23.6 | 20.2 | 2 | 1.0 | Optimal Recovery | | USR-03879 | 34 | Male | Freelance / Creative | Morning Lark | YouTube | 42 | 1 | 0 | 38 | 30.4 | 8.33 | 18.2 | 18.5 | 0 | 1.0 | Optimal Recovery | | USR-04210 | 20 | Female | Corporate 9-to-5 | Morning Lark | YouTube | 71 | 1 | 55 | 8 | 30.7 | 8.64 | 17.6 | 18.5 | 1 | 1.0 | Optimal Recovery | | USR-05021 | 38 | Female | Student | Morning Lark | Messaging / Chat | 42 | 1 | 0 | 20 | 44.0 | 7.80 | 26.4 | 19.1 | 0 | 1.0 | Optimal Recovery | | USR-02124 | 30 | Male | Freelance / Creative | Intermediate | YouTube | 33 | 1 | 0 | 32 | 84.2 | 3.66 | 20.9 | 19.8 | 7 | 9.8 | Severe Sleep Debt | | USR-04741 | 41 | Female | Corporate 9-to-5 | Night Owl | TikTok / Reels | 86 | 1 | 177 | 39 | 61.3 | 4.02 | 18.2 | 11.2 | 6 | 10.0 | Severe Sleep Debt | | USR-06860 | 37 | Female | Corporate 9-to-5 | Night Owl | YouTube | 56 | 1 | 0 | 11 | 75.0 | 3.20 | 19.8 | 18.3 | 7 | 10.0 | Severe Sleep Debt | | USR-07906 | 64 | Male | Student | Intermediate | Instagram / Reddit | 64 | 1 | 0 | 57 | 73.8 | 3.86 | 17.9 | 24.1 | 4 | 8.8 | Severe Sleep Debt | ## Category averages (all 8,500 rows) | Category | Users | % of rows | Avg sleep hrs | Avg latency (min) | Avg snoozes | Avg fatigue | |---|---|---|---|---|---|---| | Moderate Debt | 4,462 | 52.5% | 5.55 | 42.4 | 3.35 | 4.68 | | Mild Deficit | 2,004 | 23.6% | 7.23 | 34.7 | 1.81 | 1.83 | | Optimal Recovery | 1,387 | 16.3% | 8.35 | 28.9 | 0.81 | 1.08 | | Severe Sleep Debt | 647 | 7.6% | 3.76 | 72.7 | 5.88 | 9.58 | The sample rows fit these patterns. Severe-debt rows have the longest latency (61–84 min), the most snoozes (4–7) and fatigue of 8.8–10. Optimal Recovery rows have 0–2 snoozes and a fatigue score of 1.0. ## Caveats - **Weighting:** I took 4 rows per category, which over-represents the rarer categories. Severe Sleep Debt is 25% of this sample but 7.6% of the data, and Moderate Debt is 25% but 52.5%. Treat it as a tour of the row types, not as a proportional picture. - **Selection method:** I don't know how the rows within each category were picked, so I can't call them truly random. Two other pulls (15 rows each) gave similar-looking rows, but I haven't checked them for consistency beyond that. - **Column meaning:** `blue_light_filter_active` is coded 0/1, and I assume 1 means the filter was on. Units for the other columns are as named in the schema (minutes, mg, hours, percent). - **Scope:** The request was cancelled before I checked missing values, duplicates, or the distribution of occupation, chronotype and bedtime app across the full dataset.