Discover and read the best of Twitter Threads about #whew
Most recents (3)
FIRST UP from the Sea Beasties Division: No. 1 seed #1 Saber-Toothed Anchovy (Monosmilus chureloides) vs No. 2 seed Midgardia Seastar (Midgardia xandaros) #2021MMM
Tonight's Battle is a collaboration with @je_light, and I frantically double-checked some things with @DrewLab an hour ago. #2021MMM [gif is rabbit running late]
Saber-Toothed Anchovy has had a pretty uneventful tournament so far. Round 1 was easy, with a swish of the tail sending Copepod out of the field of battle & Rounds 2 & 3 ended with Pink Vent Fish & Vampire Squid meals #NomNomNom #2021MMM
UP NEXT: 2nd seed Manatee (Trichechus manatus) versus 12th seed Ring-tailed Vontsira (Galidia elegans) #2019MMM
The West Indian Manatee is a fully aquatic mammal (it never comes out of the water, like whales) with two subspecies that are found along the coast of Florida, islands of the Caribbean, & northeastern coast of South America (Husar, 1978) #2019MMM 
Truly a sea cow, manatees are herbivores & graze on all sorts of water plants. In fact, one free-ranging manatee spent on average 11 hours per day feeding (Etheridge et al, 1985) #WayToGoGus #WhereIsPickle #AwesomeOldPapers #2019MMM
Do you use diff-in-diff? Then this thread is for you.
You’re no dummy. You already know diverging trends in the pre-period can bias your results.
But I’m here to tell you about a TOTALLY DIFFERENT, SUPER SNEAKY kind of bias.
Friends, let’s talk regression to the mean. (1/N)
You’re no dummy. You already know diverging trends in the pre-period can bias your results.
But I’m here to tell you about a TOTALLY DIFFERENT, SUPER SNEAKY kind of bias.
Friends, let’s talk regression to the mean. (1/N)
Flashback to 2015. @ryan_dydx @JFBurgessJr & @jdimick1 publish a simulation paper
dx.doi.org/10.1111/1475-6… (2/N)
dx.doi.org/10.1111/1475-6… (2/N)
In a scenario that should be FINE for diff-in-diff, they get CRAZY HIGH Type I error rates.
After matching on pre-period variables (via propensity scores), things do indeed look fine. (3/N)
After matching on pre-period variables (via propensity scores), things do indeed look fine. (3/N)
