I Fooled Millions Into Thinking Chocolate Helps Weight Loss. Here's How.
John Bohannon, i09, 2015/06/01

When you hear me grumbling about the state of educational research, it is because education is second only after weight loss as the source of conflicting data, contradictory conclusions, and outright bad research. No doubt no small part of it is fabricated, though the bulk of it is no doubt created by well-meaning people. This article on weight-loss by chocolate is an example of how the system can be gamed to say, well, anything. Some low-quality research was performed, a paper was published, the press releases were written, and the rest, as they say, is history. Now this story might actually be a meta-hoax - the paper no longer exists on the journal website and the publisher denies it was ever accepted. But Retraction Watch covered it. And p-hacking is a real thing. And the news stories it generated are genuine. Via Doug Belshaw.

[Link] [Comment]

Leiden University MOOC scripts available for Download
Jasper Ginn, Leiden University, 2015/06/01

It's perhaps a bit ironic that "working with Coursera export data can be a bit daunting, especially when you are working with relatively ‘difficult’ clickstream or text data." Nonetheless, a script to export this data is now available, thanks to the University of Leiden. Jasper Ginn writes, "our git repository contains R and Python scripts that aid in the storage, extraction and pre-processing of raw Coursera MOOC data. It further contains a Vagrant virtual box which sets up a virtual environment with all required software and packages installed." Awesome.

[Link] [Comment]

The peer review drugs don’t work
Richard Smith, Times Higher Education, 2015/06/01

This is an interesting way to put it: "If peer review was a drug it would never be allowed onto the market." I'm not a great Cochrane Report supporter but I think they're right here. As the article says, peer review " it is ineffective, largely a lottery, anti-innovatory, slow, expensive, wasteful of scientific time, inefficient, easily abused, prone to bias, unable to detect fraud and irrelevant."

[Link] [Comment]