Home

My Olympic/election edition of Absolutely Maybe: Winning! Or Is It? The Science of Winners' Fate.

Back on Statistically Funny: Cupid's Lesser Known Arrow. About the growing problem of "immortal time" or "competing risk" bias.

And I updated a Statistically Funny, too. It's not often that someone does a study that basically proves one of my cartoons is real-life, not ridiculous exaggeration! I'm gonna need a wilder cartoon! On conflicts of interest: updated post.

The risks of giving research a free pass because of the prestige of the journal it's published in: my interview for journalists with Tara Haelle at the Association of Health Care Journalists (AHCJ).

And I issued another public service announcement: This is not a drill. Repeat: This is *not* a drill.

 

 

 

The day job sneaks in: I was interviewed for Retraction Watch about one of my current major projects - how post-publication activity is handled on PubMed.

Wallpapering over research weakness instead of pinpointing it: an Absolutely Maybe post on Pyschology's Meta-Analysis Problem. (If you're a systematic reviewer, keep the smelling salts handy for this one!)

I tackled another blaze of Alzheimer hype in the media at Absolutely Maybe: The Resilience of Brain Training Hype: ACTIVE Trial Redux

And added to tumblr: all about cause and effect - especially adverse effects, with the help of a dog howling at the moon (or is he?). (Alleged) effects include... Plus 2 old favorites: Disease De-Awareness Day and all about composite endpoints.

 

 

 

 

Highlight of the month for me was the wonderful annual Evidence Live conference in Oxford. I also got to give the opening keynote. A post from that: If generosity were a scientific norm... Research Nirvana: The Generosity Edition.

Another installment in my series of posts looking at the evidence and issues around social controversies on social media: Unsnarling the Complexity of Naming and Shaming.

And speaking of installments...Would you believe it's been about a year since the Tim Hunt storm? And would you believe it needed another update? So here is an update of an update.... Whatever Happened to #TimHunt?

Comment on PubMed Commons: interesting results that might not be what they seem to be, on the important issue of informed consent in IVF clinics (Canadian ones). That is my 50th comment: somehow fitting that it's a throwback to my many years as a maternity consumer advocate!

Then I also commented on a new systematic review on peer review and its limitations: my blog post on anonymity and open peer review covers a broader scope, and there was no new study found on these questions in the systematic review.

For those attending Evidence Live, I posted links to papers I talked about in the "Best Paper of All Time" session.

 

 

 

 

What was it about May and discussions about conflicts of interest everywhere? I was on that boat too: with an editorial at PLOS Medicine accompanying a new study on financial relationships between industry and organizations that develop clinical practice guidelines. And a post at PLOS Blogs: A Reader's Guide to Conflicts of Interest in Biomedicine.

Another old favorite refurbished and added to Tumblr: epidemiological melancholy over the power of the anecdote.

 

 

 

 

My first post at PLOS Blogs this month: 5 Tips to Avoid P-Value Potholes - a riveting tale of drama and statistical intrigue! My second looked at the hype around ASAPBio: Breaking Down Pros and Cons of Preprints in Biomedicine

The annual national Health Journalism conference in the US was held in a very cold and snowy Cleveland this month. On the 10th, me and Andrew M. Seaman from Reuters Health held a session on some study pitfalls and how to avoid them. I've put my slides online, together with a bunch of my blog posts and some other material in case you want to drill down on any particular issue. You can see it here at Statistically Funny.

And my first cartoon-based abstract! It's for my talk in the opening plenary for Evidence Live 2016 this June, with Ivan Oransky and John Ioannidis, and Fiona Godlee steering: Research Nirvana - The Listicle!

Speaking of understanding studies and data - which are better, natural frequencies or probabilities and percentages? A new study claims a strong answer, but I don't the contents of their paper support that conclusion. I argued why on PubMed Commons.

The month started with a profile of me by Molly Weintraub in STAT - along with a sketch portrait by Molly Ferguson.