The power imbalance between the NCAA and student athletes is appalling. Changes are happening but they're slow and long overdue. https://t.co/fIpYxyFdVR
How good is your model? I wrote (a few years ago) a step-by-step walkthrough of (part of) a Bayesian model evaluation method (Vehtari, Gelman, & Gabry, 2017) - an improvement on importance sampling in leave-one-out cross validation called PSIS-LOO 🙃 https://t.co/pa05S81HF4
Every wondered how priors and payoffs interact with perceptual confidence? Check out our now officially published paper: https://t.co/2814LA2pTc. Big thanks to my awesome collaborators @egaffincahn, @NadiaHosseini7, Pascal Mamassian, and Mike Landy!
@sarahelnozahy I like to consider two different strategies:
The "physics" approach - build the most abstract model that can account for the phenomenon.
The "kitchen sink" approach - build the least abstract model you have data to support.
Both have their uses!
@yalmoz@propublica "The consultants found it difficult to attach a dollar figure to the standards themselves" Apparently, if estimating value is hard, value can just be zero.
Excited to finally share my work on sensorimotor confidence. We examined how humans introspect about their visuomotor tracking. TL;DR: they mainly monitored tracking performance from feedback, but with a recency bias. Preprint: https://t.co/OGJqlO6Tyi
Because #Verizon doesn't have email-based customer service. Here's some feedback so other customers don't have to bang their head against a wall. @verizon@VerizonSupport@verizonfios
Sharing this resource for my fellow students at NYU (GSAS) who wish to write their PhD thesis in Latex and want a general template to start from. Comments, good or bad, are welcome.
https://t.co/yXfXLbCsIP
Our first preprint is now available! (@egaffincahn). We found confidence judgements incorporate priors and/or payoffs incorrectly: https://t.co/4ccP1XhNdr
We are excited to release our interactive map of storm surge-vulnerable sites in New York City, made in collaboration with @NYCEJAlliance.
Check it out below:
https://t.co/RUZ3r3qJvz
New @sciam article by ScAAN member @KPpipes: "Inmates in solitary confinement have far fewer protections than the nonhuman species we use in lab experiments"
https://t.co/wDIteetliq
@scipolnetwork@SEPANYC Thank you so much for the awesome symposium today! I learned a ton and it was heartening to see so much enthusiasm for science policy. #NSPNsymposium18