Measuring the predictability of life outcomes with a scientific mass collaboration.
Matthew J SalganikIan LundbergAlexander T KindelCaitlin E AhearnKhaled Al-GhoneimAbdullah AlmaatouqDrew M AltschulJennie E BrandNicole Bohme CarnegieRyan James ComptonDebanjan DattaThomas DavidsonAnna FilippovaConnor GilroyBrian J GoodeEaman JahaniRidhi KashyapAntje KirchnerStephen McKayAllison C MorganAlex PentlandKivan PolimisLouis RaesDaniel E RigobonClaudia V RobertsDiana M StanescuYoshihiko SuharaAdaner UsmaniErik H WangMuna AdemAbdulla AlhajriBedoor AlShebliRedwane AminRyan B AmosLisa P ArgyleLivia Baer-BositisMoritz BüchiBo-Ryehn ChungWilliam EggertGregory FalettoZhilin FanJeremy FreeseTejomay GadgilJosh GagnéYue GaoAndrew Halpern-MannersSonia P HashimSonia HausenGuanhua HeKimberly HigueraBernie HoganIlana M HorwitzLisa M HummelNaman JainKun JinDavid JurgensPatrick KaminskiAreg KarapetyanE H KimBen LeizmanNaijia LiuMalte MöserAndrew E MackMayank MahajanNoah MandellHelge MarahrensDiana Mercado-GarciaViola MoczKatariina Mueller-GastellAhmed MusseQiankun NiuWilliam NowakHamidreza OmidvarAndrew OrKaren OuyangKaty M PintoEthan PorterKristin E PorterCrystal QianTamkinat RaufAnahit SargsyanThomas SchaffnerLandon SchnabelBryan SchonfeldBen SenderJonathan D TangEmma TsurkovAustin van LoonOnur VarolXiafei WangZhi WangJulia WangFlora WangSamantha WeissmanKirstie WhitakerMaria K WoltersWei Lee WoonJames WuCatherine WuKengran YangJingwen YinBingyu ZhaoChenyun ZhuJeanne Brooks-GunnBarbara E EngelhardtMoritz HardtDean KnoxKaren LevyArvind NarayananBrandon M StewartDuncan J WattsSara McLanahanPublished in: Proceedings of the National Academy of Sciences of the United States of America (2020)
How predictable are life trajectories? We investigated this question with a scientific mass collaboration using the common task method; 160 teams built predictive models for six life outcomes using data from the Fragile Families and Child Wellbeing Study, a high-quality birth cohort study. Despite using a rich dataset and applying machine-learning methods optimized for prediction, the best predictions were not very accurate and were only slightly better than those from a simple benchmark model. Within each outcome, prediction error was strongly associated with the family being predicted and weakly associated with the technique used to generate the prediction. Overall, these results suggest practical limits to the predictability of life outcomes in some settings and illustrate the value of mass collaborations in the social sciences.