Elizabeth Stuart
@lizstuart.bsky.social
Statistician; Professor and Chair @JHUBiostat @JohnsHopkinsSPH, w/links to @SREESociety, @AmericanHealth. Oh, & spouse, mom, runner, traveler.
Thank you! It's such an honor, and a wonderful prompt to reflect on how thankful I am to have found this career!
October 23, 2025 at 3:54 PM
Thank you! It's such an honor, and a wonderful prompt to reflect on how thankful I am to have found this career!
This is SO wonderful and well deserved; congratulations!!
October 8, 2025 at 5:46 PM
This is SO wonderful and well deserved; congratulations!!
Yes, fantastic messages -- I'm glad you wrote all of this down!
August 14, 2025 at 1:39 AM
Yes, fantastic messages -- I'm glad you wrote all of this down!
I was SO sad that my own session conflicted with this one!
June 11, 2025 at 8:56 PM
I was SO sad that my own session conflicted with this one!
My daughter was taking the same exam today! I kind of love the coordinated timing all around the country and the idea of a bunch of kids all sitting down for the tests in a coordinated timing kind of way.
May 7, 2025 at 8:53 PM
My daughter was taking the same exam today! I kind of love the coordinated timing all around the country and the idea of a bunch of kids all sitting down for the tests in a coordinated timing kind of way.
“Scale is not a substitute for scrutiny” may be my new favorite quote. Thanks @adamjkucharski.bsky.social!
April 25, 2025 at 10:05 AM
“Scale is not a substitute for scrutiny” may be my new favorite quote. Thanks @adamjkucharski.bsky.social!
Make sure it's done multiple times -- otherwise imputation will make estimates appear more precise than they really are.
February 9, 2025 at 9:53 PM
Make sure it's done multiple times -- otherwise imputation will make estimates appear more precise than they really are.
One add'l insight is that in some contexts a "missing data indicator" approach is actually okay (that's what twang does in implementation), and even if you do MICE [with outcomes and treatment included too!] it might make sense to include missing data indicators, as they may carry info about people.
January 28, 2025 at 12:15 AM
One add'l insight is that in some contexts a "missing data indicator" approach is actually okay (that's what twang does in implementation), and even if you do MICE [with outcomes and treatment included too!] it might make sense to include missing data indicators, as they may carry info about people.