Do Dating Apps Have a 6'0" Problem? It's... Complicated...

Get Nebula using my link for 50% off an annual subscription: https://go.nebula.tv/notdavid Accompanying behind the scenes video on Nebula: https://nebula.tv/videos/notdavid-why-did-the-heights-video-take-so-long-post-commentary?ref=notdavid Music and sources below! *CORRECTIONS/ERRORS*: 1) I made a really dumb error with visualizing the metric heights. By using 2.5cm bins the number of integers can change from three-to-two between bins. This is only for the visualization, there is still overreporting at every 5cm at the 1cm level so I think the conlcusion still holds, but the barchart visually skews this. Thank you to the viewers who pointed this out and I'm sorry about this mistake. 2) At 4:23, should read something closer to 40% is the chance to see 3489 men, OR MORE. BlueSky: https://bsky.app/profile/notdavidyt.bsky.social Insta: https://www.instagram.com/not.david.yt/ If you'd like to support these videos: https://www.patreon.com/NotDavid Thumbnail credit: One of the AB tested thumbnails (Expectations/Reality) was inspired by a Stand Up Maths thumbnail (We have statistical evidence that people are mildly psychic). Notes: Testing the effect of orientation - we want to distinguish if the 4700 nonstraight men actually follow a different distribution of heights, or if we are just seeing chance differences. The idea is that we'll take all 35000 men, and then randomly select 4700 men. Then, we calculate how "close" is this random sample of men's heights to the original 4700. This "closesness", can be, if you're not used to it abstract but in the same way you and I can be some distance apart, statisticians can define how close two barcharts/histograms are. One such distance is the Kolmogorov-Smirnov (KS) distance, which is the one I used. So we get that distance, then do it again with a new random sample, over and over. If those distances are typically very small then we can say that the heights of the 4700 nonstraight men were representative of the whole population, and so it is possible they are equivalent to just randomly taking men. But if the distances are high, then we can say it is not likely that our 4700 men are statistically similar to a random sample of 4700 men. Sources in order: [1] Main paper and data: OkCupid Data for Introductory Statistics and Data Science Courses, A. Y. Kim, and A. Escobedo-Land [2] Partner height preferences paper: The height of choosiness: mutual mate choice for stature results in suboptimal pair formation for both sexes Stulp G. et al. [3] Counter example: Height preferences in humans may not be universal: Evidence from the Datoga people of Tanzania. P. Sorokowski, and M.L. Butovskaya [4] 400K self-reported heights data: The Behavioral Risk Factor Surveillance System (BRFSS), U.S. C.D.C. [5] Measured heights data: National Health and Nutrition Examination Survey, U.S. C.D.C. [6] Supplementary measured heights from 2015-2018 (agregated stats no individual data points, I used tables 10 and 12): Anthropometric Reference Data for Children and Adults: United States, 2015-2018, U.S. C.D.C. [7] Vox Article on OkCupid profile leak: Researchers just released profile data on 70,000 OkCupid users without permission, B. Resnick [8] German height statistics (I used table 1): Anthropometrische Messungen in der NAKO Gesundheitsstudie – mehr als nur Größe und Gewicht. B. Fischer et al. [9] Self reported over-estimates may vary by country: Self-reported and measured weight, height and body mass index (BMI) in Italy, the Netherlands and North America. A. J. Krul, H. A. M. Daanen H. Choi [10] How self-reported end-digit bias' can negatively effect health metrics: Terminal digit bias in a specialty hypertension faculty practice. S. Thavarajah, W. B. White, G. A. Mansoor [11] How rounding in medical devices may also negatively effect health metrics: End-digit preference in general practice: a comparison of the conventional auscultatory and electronic oscillometric methods. M. Burnier, U. E. Gasser Music (in order, if source not stated then it is from youtube library): 1. Identical Honeymooners, Chris Doerksen (https://chrisdoerksen.bandcamp.com/album/december-5th-2024) 2. Mas Cafe, Casa Rosa 3. A Night Alone, Track Tribe 4. Sightseeing, Chris Doerksen (https://chrisdoerksen.bandcamp.com/track/sightseeing) 5. Mediterrarean View, Everet Almond 6. Spring Field, Godmode 7. Test Demo Bass, Alge 8. Event Horizon, World Complete (https://worldcomplete.bandcamp.com/album/holograms-album-remastered) 9. Loading Screen, Dyalla 10. RPG Store, Chris Doeksen (https://chrisdoerksen.bandcamp.com/album/looking-for-light) 11. Fate (mistakenly called the Fool's Game in video, that is the album), Seimuc (https://open.spotify.com/artist/1kQxwXIPOIZKYCZbwIw4Pz?si=e97873f647cb46a6) 12. Sun Machine One, Loopop 13. Intermission, Vaporwaves + Stux.Io (https://stuxio.bandcamp.com) 14. Sure Thing, Chris Doerksen (https://chrisdoerksen.bandcamp.com/album/distance) 15. Wehrmut, Godmode