I was just reading this article about Statscan, the Canadian warehouse for storing data and the branch of the government charged with the statistical analysis of all things Canadian, and came across this dizzyingly amazing quote about Statscan:
Its statistics are so routinely cited as fact by all parties, factions and interests that “Statscan says” is effectively a Canadianism meaning “the following is true.” It’s hard to overstate how valuable that makes this dull little agency.
It made me wonder, does the equivalent statement exists in the US? Has anyone ever said the same thing about the Bureau of Labor Statistics? The Census Bureau? Any fact ever?
We’ve posted about stats blogs here before. Evelyn first wrote about some of her favorite stats blogs in 2013 with a reprise in 2015 and a post dedicated to the much maligned — although actually useful when used correctly and honestly — p-value. Given that some time has passed, I though we were due for a reprise to the reprise. So here we go.
I was recently reminded of the statistician, political scientist and blogger Andrew Gelman when he posted an essay on the replication crisis to the special science at 40 series for the New York Times. Gelman co-writes a blog about statistical modeling, causal inference, and social science. On the blog Gelman addresses statistical methods and then some of the not-so-obvious subtleties of election statistics and voter turnout, and a whole bunch of other statistically infused and politically adjacent topics.
Another fantastic stats blog is StatChat which is hosted out of the University of Auckland in New Zealand. I especially love blogger and biostatistician Thomas Lumley’s Briefly posts which give a roundup of recent interesting stats objects from the week, which is where I found the Statscan article that started all of this.
You should also check out Simply Statistics, a great blog written by three biostatisticians. I really enjoyed their recent post on the intricacies of gathering data for their analysis on mortality rates from the hurricane in Puerto Rico. On the slightly more technical end, I also recommend Frank Harrell’s blog Statistical Thinking.
Given my personal interests, I tend to lean towards statistics practiced in the service of social science. But if clinical and biomedical research statistics are the kind of thing that you think about in your quiet moments, then this clinical research statistical methodology free association word game — now there’s a string of words I never thought I would use in succession! — from the Swiss Medical Weekly is made for you. And if I ever order a Bud Light when Westmalle is on tap I will be sure to phone in the order to a bartender at a different bar down the street, ask him to serve it to a patron there, and have that patron call my landlord to report on the quality of the Bud Light.
Tripel blind. Get it? I’ll see myself out.
Are there any stats blogs that you really love? Have you ever heard an Americanism giving emphatic support of the Census Bureau? If so, let me know on Twitter @extremefriday.
PS on Monday November 26th the AMS is celebrating 130 years with #AMSday and offering discounts on AMS and MAA Press books, free access to mathscinet, and special perks for renewing (or starting!) an AMS membership.
One long-running one I’d definitely add to those you’ve already mentioned is Virginia Mayo’s blog:
It can be a bit technical or philosophical but she often has interesting, extended discussions with Frank Harrell, Stephen Senn, Gelman, & others.