Profile cover photo
Profile photo
David Richter
322 followers
322 followers
About
David's interests
View all
David's posts

Post has shared content
I second this.
Dear Everyone Who Will Be Hurt More By This Than I Am,

You steer. I'll row. I am not leaving you here.

Love you all,
Andy

Are you eligible to vote in the US elections? Then please do.

Please vote against Trump.

Please vote the whole ballot. There are lots of critical races other than the presidency.

Think you are eligible but not registered? In 12 states and Washington DC, you can register in person at the polls. http://www.ncsl.org/research/elections-and-campaigns/same-day-registration.aspx

Not sure where to vote? Try this: https://g.co/kgs/NuZTch

#IVoted

Post has attachment
Tea for thirsty nerds! Inclusive + family-friendly game/craft nights + book clubs! Local art + ethically sourced ingredients! What's not to love? Check out Friday Afternoon's Kickstarter and consider backing them to get some awesome teas and exclusive invitations when they expand into a cool brick-and-mortar tea shop in Seattle.

Post has attachment
I got out to see Kubo last night. It's a really excellent movie. Don't believe me? Rotten Tomatoes has it at 97% fresh [1].

This track is the first one in the credit role. It's a very good cover.

[1] https://www.rottentomatoes.com/m/kubo_and_the_two_strings_2016/

Post has shared content
Horrifying, really, and will hopefully spur a transition to open data for all published clinical trials of approved or up-for-approval drugs.
Clinical trials are formally high-quality controlled trials, but the raw data is a trade secret of the drug company -- not even the peer reviewers see it. Here a court released some data and documents, so we get a glimpse of what can be hiding: the primary statistical significance was massaged, its effect size was egregiously misstated (as d = 2.9 instead of 0.32!), unfavorable metrics were dropped, and a new metric was tailored onto the data to give a rosier statement.

The drug company wanted to market Celexa for children and adolescents, and they got a published peer-reviewed randomized trial saying it was safe and effective. It actually failed to show the expected difference from placebo, especially in children, but doctors gave it to children trusting this study.

Why so dodgy? Probably related to the fact that the paper was written by a communications consultant hired by the drug company. The academic 'author' was chosen and sent their full draft. Email said they wanted to ghostwrite this paper for "speed and greater control", control needed because "I've heard through the grapevine that not all the data look as great as the primary outcome data". These secondary metrics were dropped from the paper, and they then chose a paper format saying "As a Brief Report, we feel we can avoid mentioning the lack of statistically significant positive effects at week 8 ..." Not even most commercial ghostwriting would be as sketchy as this, I would hope, but the entire fact of ghostwriting was then not disclosed with the published paper.

http://content.iospress.com/articles/international-journal-of-risk-and-safety-in-medicine/jrs717

Post has shared content
Me: "I dunno. I'm not 'with her.' My politics can't be summed up in a slogan that fits on a T-shirt."

My brother: [shows picture of T-shirt]

Me: Disregard, then.

Post has shared content

Post has shared content
A frightening analysis of Trump as demagogue and would-be tyrant.

Post has shared content

Post has shared content
Wait while more posts are being loaded