October 30, 2016
Briefly
- A long post on the use and misuse of the ‘Twitter firehose’, from Bloomberg View
- A long story at Stuff about discharge without conviction, though a bit undermined by the fact that, as the story says, “[the] number of discharges without conviction has plummeted, from 3189 in 2011, to 2103 in 2015,“
- While the idea of predicting the US election using mako sharks (carchariamancy?) is no sillier than psychic meerkats or lucky lotto retailers, I don’t think the story really works unless the people pushing it at least pretend to believe it.
- On the other hand, some people did seriously argue that shark attacks affected the results of presidential elections. And were wrong
- There’s been a lot of interest in the voting intentions of evangelical Christians in the US for this election: they’re a moderately large and active group. They’re also hard to identify in polling. Ed Setzer writes about a new definition. Fred Clark explains why the definition is hard — partly because of issues the new definition wants to address, but also because of issues it’s trying to bypass.
Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »