dougthompson

@CAndersonStats @RoyalStatSoc Great to see this Craig! Well done!

It's an honour to have been selected by @RoyalStatSoc as one of the William Guy Lecturers for the coming year. I absolutely love sharing my passion for our wonderful subject with young people. Hopefully this can inspire and excite them about statistics. twitter.com/RoyalStatSoc/s…

Proud to see this Special Issue coming out with 10 papers on statistical challenges in clinical trials for COVID-19 treatments, vaccines, and diagnostics. Thanks to Freda Cooner, @ToshimitsuHama1, Gene Pennello and @posch_m for the great collaboration! tandfonline.com/toc/usbr20/14/1

@AdmiralFallow @ChemUnderground @Bandcamp Just wow! This album is nothing short of beautiful!

The Idea Of You is out now! 5 years in the making, we're delighted to finally share this record with you all. Go and seek it out wherever you discover new music. Thank you very much for your support. AFx smarturl.it/AdFalTheIdeaOf… @ChemUnderground @Bandcamp #BandcampFriday

@LopezAyalaP @tmorris_mrc @f2harrell This simulation study might be an interesting read? tinyurl.com/8687kfpc

(2/2)Stratified randomization or minimization done. Analysis should reflect design of the RCT & stratification/minimisation Vars should be adjusted for in the analysis BUT few events. What to do? Adjust for all necessary covariates despite not enough d.f.? @tmorris_mrc @f2harrell

(1/2)The benefits of prognostic covariate adjustment in moderate/large sample size RCT is clear (⬆️ in power, protection against chance imbalance).How about small sample size RCT (<50 per arm)? Suppose we have 20 events, still possible to adjust?How many covariates? Issues w d.f?

@JennaJrdn @Randy_Au Some refs fharrell.com/post/medml/ & tinyurl.com/jmcayt9u

@JennaJrdn @Randy_Au There is no such golden rule. Patient data is noisy so parameters to estimate such individual effect need to be treated with caution. ML does not get around this; ML models can & do suffer from poor calibration. Also stats can say things about the individual, see eg n-of-1 trials

@Drkjegan Gorgeous!!

Another fine, crisp morning over #glasgow this morning 😁

Thrilled about our fantastic featured speakers for R/Pharma 2020 - Christina Fillmore (GlaxoSmithKline), @topepos (RStudio) and @Viral_B_Shah (Julia Computing)! rinpharma.com/speaker/ #rinpharma

@Drkjegan Well done Kieren! Very well deserved

Just received confirmation of my promotion to Research Fellow at Strathclyde. Really happy to learn this, and I'm looking forward to continuing to build collaborations with exceptional people both near and far! #digitalhealth #collaboration #teamwork

My take on communicating about coronavirus: medium.com/wintoncentre/c…

@graemearcher Happy Birthday!!

@TPS_Tree Red

@TPS_Tree Bisque

@TPS_Tree Yellow

@TPS_Tree Indigo

@MikeKSmith Thanks for sharing this! There’s a ton of wee gems in here

ggplot2 tips, tricks and a quick scamper through some extension packages. rpubs.com/MikeKSmith/551…

Gratified that >1500 people interested in biomedical research have signed up for the BBR course starting Oct 4 at 10am EDT. Registrants will receive access information by Sunday. Not too late to register - hbiostat.org/bbr #bbrcourse

Heading home after a fascinating one day meeting on how ML/AI is being used in pharma. A key take away for me is recognising when ML/AI can offer insight that established statistical methods don’t already cover. #PSIstats #AI #ML

Biostatistics for Biomedical Research almost-weekly web course registration is now open. Go to hbiostat.org/bbr for course details and registration link. @EdgeforScholars #bbrcourse @vandy_biostat @VUMChealth #StatThink

Toying with the idea of hosting an almost-weekly one-hour live webinar (with student participation) on applied statistics/biostatistics. Some topic choices would come from twitter polls. Please respond to the poll in the next tweet if you are definitely interested.