Dr. Rachel Sippy is on STRIKE!

Proud of this joint effort with @ElizabetLilia @marobertvt @DrAnnaStewart and collaborators!

@GermsAndNumbers Just quietly panicking that my data might be useless and that the new collection methods we've worked on are now obsolete

@eeid2023 White-tailed deer in an urban neighborhood of Pittsburgh - I was shocked to find her and immediately thought of @swientist SARS-CoV-2 transmission research (I did not cough on her) @MegganCraft #eeid2023

Can relate

Highlighting disease forecasting challenges by @alexvespi and ending on a critical note - the old model of modeling is out! #eeid2023

Deciding to say hello at a conference: have I actually met this person or do I have a parasocial relationship with them because of Twitter? 🤔 #eeid2023

@HScottFordsmand @clarehall_cam ^actual footage of me studying ML

Feel like a have a slightly better grasp if machine learning and AI after today’s lunch session w/ @GermCurves at the @clarehall_cam HPS Special Interest Group on ‘Machine Learning and Prediction in Clinical Medicine’.🤖🩺 Took me a while to get all the 🌳/forest refs though 🤦‍♀️

@EpiEllie I was just having a discussion about this - evidence can sometimes be context-specific! As you've said, we need to focus on what we are trying to achieve, and consider whether the "evidence" we'd like to apply is relevant or even "evidence" at all

“Evidence-based” proponents often get very angry when you suggest the “evidence” they’re using isn’t right or isn’t saying what they think it is. Which, idk about you, but seems to me that if they really cared about evidence, they would be extra interested in evidence quality🤷🏼‍♀️

The thing about “evidence-based” medicine or parenting or nutrition or whatever is that people who hype this approach never seem to care to much about how that “evidence” is obtained and whether it can actually inform any of the choices they want to make.

@GermsAndNumbers @MattMietchen Well done Dr Mietchen!!

Lots of people have preconceived ideas about what working in a university must be like. Here, UCU member Chris has a message about the reality of it for far too many staff. Share the video. Support the strike. Join us in demanding change. #ucuRISING

@lyzl Fun fact: the health benefit of a heterosexual marriage is only seen for the man! Men have this health benefit, women do not.

@MaartenvSmeden Nice comment by @eodjidja thelancet.com/journals/langl…

Today, university staff go on strike They are striking for decent pay, pensions and secure employment They are striking for students They are striking for the future of higher education RT IF YOU BACK THEM #ucuRISING

@avsmph @jfeldman_epi For me, the RE is something that makes people more similar to each other, usually in a way that's hard to define (and may not be captured well by other covariates in the model). So then the RE could be insurance coverage, households, pts with the same physician, etc. 2/2

@avsmph @jfeldman_epi I think it helps to think of multilevel/hierarchical models as a type of mixed effects model (model with RE), and then it's easier for the concept of an RE to be applied to many things. 1/2

Putting this as a quote in a lecture