May 30, 2015

Briefly

  • Michael LaCour, the researcher accused of faking data in his study on changing opinions about same-sex marriage, has issued a response to the allegations today as he promised. Virginia Hughes at Buzzfeed and  Ivan Oransky at RetractionWatch have stories. I’ve seen a lot of comments about this on Twitter from people who work in empirical social sciences or in ethics. So far I haven’t seen anyone who is convinced by the response.
  • There’s story in New York magazine about David Broockman, who found the problems, about the difficulty of reporting suspicions of fraud in research.  ““I think there’s an interesting metaphor between what I went through now and what I went through as a gay teenager,” Broockman says.
  • The UK is introducing a ‘legal highs’ ban. The government are using a widelyquoted figure of 97 deaths in the last year due to legal highs.  As Vice.com explains, the figure is bogus in two ways. Firstly, the figure is based on detectable presence of a drug, not on it being even a contributing cause. Secondly and more importantly, the drugs in the majority of these cases were already illegal at the time.
  • You’ve probably seen the Washington Post chart of deaths in World Cup construction. Unless you saw it at the original site, you may not have seen the disclaimer above it (emphasis added):

Some of these numbers (like Sochi’s) are third-party estimates, others (like Beijing’s) are based on official numbers that are almost certainly an undercount. And it’s tough to do an apples-to-apples comparison here, since the Qatar estimates include the deaths of all migrant workers after the announcement of Qatar’s successful bid in 2010, while other countries’ figures may only include deaths directly related to, say, stadium construction.

The exploitation of migrant workers isn’t new and it isn’t all Sepp Blatter’s fault. It’s one of the reasons Qatar was regarded as a terrible choice at the time.
Social media tends to spread images out of context, and this can be a problem if you’re trying to be seen as the sort of journalist who cares about facts.

avatar

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 »