I haven’t read the paper but above-chance accuracy doesn’t seem implausible to me. I’m curious if you think it’s implausible on its face or if you see problems in the paper (or something else)? (Full disclosure, I’ve done similar work: https://journals.sagepub.com/doi/abs/10.1177/0146167209346309 …)
-
-
-
I responded to that post with critique about the high false positive rate for that method (~ 90% for 2-tailed tests) and that their effective N is 6, no matter how many rater replications they have. Looking at the images and method, it’s actually 2 (the 3 traits have big overlap)
- Još 12 drugih odgovora
Novi razgovor -
-
-
Wait they think you can tell personality from faces?
-
Not mine. I'm an inscrutable non-white person.
- Još 2 druga odgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.