Unknown

Dataset Information

0

Measuring the predictability of life outcomes with a scientific mass collaboration.


ABSTRACT: 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.

SUBMITTER: Salganik MJ 

PROVIDER: S-EPMC7165437 | biostudies-literature | 2020 Apr

REPOSITORIES: biostudies-literature

altmetric image

Publications

Measuring the predictability of life outcomes with a scientific mass collaboration.

Salganik Matthew J MJ   Lundberg Ian I   Kindel Alexander T AT   Ahearn Caitlin E CE   Al-Ghoneim Khaled K   Almaatouq Abdullah A   Altschul Drew M DM   Brand Jennie E JE   Carnegie Nicole Bohme NB   Compton Ryan James RJ   Datta Debanjan D   Davidson Thomas T   Filippova Anna A   Gilroy Connor C   Goode Brian J BJ   Jahani Eaman E   Kashyap Ridhi R   Kirchner Antje A   McKay Stephen S   Morgan Allison C AC   Pentland Alex A   Polimis Kivan K   Raes Louis L   Rigobon Daniel E DE   Roberts Claudia V CV   Stanescu Diana M DM   Suhara Yoshihiko Y   Usmani Adaner A   Wang Erik H EH   Adem Muna M   Alhajri Abdulla A   AlShebli Bedoor B   Amin Redwane R   Amos Ryan B RB   Argyle Lisa P LP   Baer-Bositis Livia L   Büchi Moritz M   Chung Bo-Ryehn BR   Eggert William W   Faletto Gregory G   Fan Zhilin Z   Freese Jeremy J   Gadgil Tejomay T   Gagné Josh J   Gao Yue Y   Halpern-Manners Andrew A   Hashim Sonia P SP   Hausen Sonia S   He Guanhua G   Higuera Kimberly K   Hogan Bernie B   Horwitz Ilana M IM   Hummel Lisa M LM   Jain Naman N   Jin Kun K   Jurgens David D   Kaminski Patrick P   Karapetyan Areg A   Kim E H EH   Leizman Ben B   Liu Naijia N   Möser Malte M   Mack Andrew E AE   Mahajan Mayank M   Mandell Noah N   Marahrens Helge H   Mercado-Garcia Diana D   Mocz Viola V   Mueller-Gastell Katariina K   Musse Ahmed A   Niu Qiankun Q   Nowak William W   Omidvar Hamidreza H   Or Andrew A   Ouyang Karen K   Pinto Katy M KM   Porter Ethan E   Porter Kristin E KE   Qian Crystal C   Rauf Tamkinat T   Sargsyan Anahit A   Schaffner Thomas T   Schnabel Landon L   Schonfeld Bryan B   Sender Ben B   Tang Jonathan D JD   Tsurkov Emma E   van Loon Austin A   Varol Onur O   Wang Xiafei X   Wang Zhi Z   Wang Julia J   Wang Flora F   Weissman Samantha S   Whitaker Kirstie K   Wolters Maria K MK   Woon Wei Lee WL   Wu James J   Wu Catherine C   Yang Kengran K   Yin Jingwen J   Zhao Bingyu B   Zhu Chenyun C   Brooks-Gunn Jeanne J   Engelhardt Barbara E BE   Hardt Moritz M   Knox Dean D   Levy Karen K   Narayanan Arvind A   Stewart Brandon M BM   Watts Duncan J DJ   McLanahan Sara S  

Proceedings of the National Academy of Sciences of the United States of America 20200330 15


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.  ...[more]

Similar Datasets

| S-EPMC9312390 | biostudies-literature
| S-EPMC8297818 | biostudies-literature
| S-EPMC3683734 | biostudies-literature
| S-EPMC6279741 | biostudies-literature
| S-EPMC4391948 | biostudies-literature
| S-EPMC4299220 | biostudies-literature
| S-EPMC9401607 | biostudies-literature
| S-EPMC7211484 | biostudies-literature
| S-EPMC3980335 | biostudies-other
| S-EPMC9844855 | biostudies-literature