A few years ago, Susan Greenfield said something zany, even for Susan Greenfield: that the rise in autism diagnoses were related to the Internet.

As Robery Olley once noted, she is the scholar who in 1994 during her famed Royal Institution Christmas Lecture claimed that playing Prokofiev at half speed was linked to suicide. That was not a shock to music listeners but it was to serious scholars.

Her bold nonsensical claims did not end there. Despite having 30 honorary degrees, The Royal Institution tried to get rid of her and her zany, expensive ways and finally did in 2010. She claimed it was sexual discrimination, which is always an easy, subjective charge to level, while they said they just wanted to get rid of her contract. And her reputation. In a Kardashian Index way, she has always ranked quite high, but her actual science has been a head scratcher.

By 2011 she went from gradually zanier regional media talking head to full on international nonsense spouter, claiming that the Internet was leading to an autism epidemic. Her basis was a PLOS One paper stating that peoples' brains change. 

After the backlash, she claimed "I point to the increase in autism and I point to internet use. That's all." This really sent the Internet into fits, because it was shockingly irresponsible for a media figure. Science journalist Carl Zimmer even created the #greenfieldism meme.
But now she is back with a whole book stuffed full of the kinds of spurisous correlations she is noted for. And Dr. Dorothy ("Deevy) Bishop, Professor of Developmental Neuropsychology at Oxford, notes that the sources for her claims haven't gotten any better than pointing to Internet use and increases in autism. The Daily Mail, grey literature. In a world where there are 25,000 journals even Greenfield's "peer reviewed" literature claims are meaningless, especially if they are all epidemiology correlations that she picked because it had the conclusions she wanted.

One example Bishop notes is a 2009 paper which doesn't even mention the Internet.  It's epidemiology, which is a red flag in the modern era, and I would mention that I recently debunked another paper by those authors, Delwiche and Hertz-Picciotto, claiming that farms caused autism without ever actually doing any science to show it - which is the same thing Greenfield does.  Hertz-Picciotto is on the Advisory Board of Autism Speaks and on the board of an anti-chemical advocacy group and she got paid by Autism Speaks to write that paper, which she did not disclose as a conflict of interest. Regardless, anyone looking for a way to make a claim about autism is going to find a Hertz-Picciotto paper they like.

Bishop goes into detail about the problems with all of Greenfield's source material, so there is no need to rehash it in detail here. Greenfield claims 500 peer-reviewed papers, Bishop notes there are actually 395 sources and some of them are just bizarre, like statistics on tobacco use and life expectancy. And I like the Daily Mail, it makes me laugh, but I wouldn't use them as a science source. Believe me, I tried in Science Left Behind, but my fact checker did not agree that my desire for humor merited including the Daily Mail.

Greenfield does, though.