Everything Hertz

90: Mo data mo problems

Informações:

Sinopse

Dan and James discuss two listener questions on performing secondary data analysis and the potential for prestige to creep into open science reforms. More info and links: Why generate your own dataset when you can get a high impact paper using public data? Thanks to Stu Murray (https://twitter.com/DrStuartBMurray) for the question Will people steal your ideas? The journal Scientific Data (https://www.nature.com/sdata/) Are we now incentivising data mining rather than data collecting? Synthetic data Dan’s recent synthetic data preprint primer (https://psyarxiv.com/dmfb3/) Ego and prestige got us into the mess we’re trying to fix with open science, but how can we stop this from happening again? Thanks to Robin Kok (https://twitter.com/robinnkok) for the question, listen to our episode with him on e-health (https://everythinghertz.com/34)! Did all the people who co-authored the paper to change statistical significance the default p-value threshold to .005 (https://www.nature.com/articles/s41562-017-0189-z) actu