Notice my new 🏳️🌈🏳️⚧️📊 banner? I’ve updated my #rstats pride flag code to use the wonderful color palettes in the ‘gglgbtq’ package (by @Rturtletopia)! Every generated plot is unique and special, make and share yours today! https://t.co/KR9Oo5vpbh
I know it's pretty dead on here, but my team (not me) within the data & computational science team at Vertex is hiring an integrative data scientist: https://t.co/SBpQUY5zjS. If you're interested please apply, it's an awesome place to work!
Ever wonder how many lives have been saved by NIH-funded research - including your own? Enter any medical condition and instantly see how your tax dollars transformed science into survival.
https://t.co/ukj6bhL1tW
*BROOKLYN*, I’ll be back on Thursday 5/1 at 10am, where it all began at Kings County Hospital where my mother and grandmother worked, where I trained in Emergency Medicine, to give a keynote!
The first 100 attendees will receive a free copy of LEGACY🩺
@5_utr@ADAlthousePhD@f2harrell Agreed. Establishing a clear target population, identifying relevant characteristics, and having an established research question can help shift ‘representativeness’ from a sweeping concern to something more concrete!
@ADAlthousePhD@5_utr@f2harrell Also good to consider what types of questions these tools can help best answer, like using trial results (across diff diseases) to inform population-level decisions (e.g. policy, reimbursement). Reconciling any trial/target pop differences can help build confidence!
@ADAlthousePhD@5_utr@f2harrell There are ways to improve representativeness from a design perspective (like broader eligibility criteria, thoughtful recruitment/sampling). These stat methods can help when such design strategies are infeasible, or also if trials have already recruited/read out
📚 Take a break from doomscrolling and pick up a book from your local library or support an independent bookstore! Here are some of my favorite novels from last year: