Like a good sharknado, Susan Greenfield is (a) ridiculous and (b) back for more.

We all remember this defence of her claim that internet use causes autism, don’t we?

I point to the increase in autism and I point to internet use. That is all.

Well, whoopy do. On that basis, Russia’s annexation of Crimea was responsible for loom bands. Obviously.

As a reminder, Greenfield’s schtick is as follows: according to her, social networks — the internet kind — cause brain damage. Now such phrasing sounds like a jokey summary of something more nuanced. However, it’s pretty much everything in a nutshell.

But while Greenfield is an expert in treatments for Parkinson’s and Alzheimer’s disease, she has no expertise or training in sociocultural factors that actually cause brain damage. In other words, she knows X but doesn’t know Y. She knows how to get the milk into the tea, but it doesn’t logically follow that she can get it back out again.

Also, she has no expertise or training in autism. Or in developmental psychology more generally. Or in psychological assessment and diagnosis. Or for that matter, in internet behaviour, sociology, engineering, or any relevant field.

Here’s a general tip for all you logic fans out there: knowing a lot about X doesn’t mean that you’ll know anything at all about Y. Alan Hansen knows a lot about where right-backs should stand when defending set-pieces in football. However, I wouldn’t rely on him to flash a custom ROM onto my Xperia ZL.

Critics have been vocal in pointing out that Greenfield has not actually chosen to submit her theories to the standard due-diligence practice used in professional science, namely, peer-review. In short, she hasn’t submitted her work to a journal. Yes, she’s made speeches. Yes, she’s given interviews. Now she’s written a book. All of the stuff that philosophers of science sarcastically refer to as “science by press conference.”

But she simply hasn’t presented her theory for formal evaluation by her fellow scientists. This means that when problems are pointed out, she can just claim she’s been misquoted, or that she didn’t really mean what it sounded like she said. When she is accused of making erroneously disparaging claims about autism, she can argue that her claims were taken out of context.

For example, in today’s Telegraph interview about her book, the interviewer points out that one of the studies she cites was published in 2004, before either Facebook or Twitter even existed. Like any good shyster, Greenfield responds not by changing her belief, but instead by massaging her claim. The journalist paraphrases her response as follows:

the existence of email could have had a similar effect.

So you see? Now it’s email that damages the brain. Soon it will be talking. Talking damages the brain.

In her book, Greenfield provides hundreds of references to studies that she says are relevant to her claims. Here’s how she formulates her conclusions:

If on balance it comes through that there are more negative ones than positive ones, then I think we can only conclude that there are more negative ones than positive ones.

Tautology notwithstanding, Greenfield’s basic reasoning is that if more studies show one effect than another, then it’s more likely that effect exists. Let’s not bother ourselves to think about study size, study quality, research method, or editorial patterns that favour the publication of controversial claims. In Greenfield’s world, two bad studies outweigh one good one.

That’s assuming there is a good one. When the interviewer suggests that matters could be resolved by a study of 2,000 people, Greenfield responds:

How do you get 2,000 people, [or] the amount of money that would take…?

That’s right folks. Because no study ever has had 2,000 people. It’s just not possible. So quit yer complaining about the small studies, right?

Here’s a short list of what makes an activity pseudoscientific. See how many apply in the present case:

• Purporting to be scientific but failing to adhere to the standard rigours of scientific methodology
• Science by press conference/avoidance of peer review
• Claims of omnicompetence
• Hypotheses that can’t be falsified
• Apocalyptic reasoning
• Appeals to holism
• Vague definitions of terms (e.g., “Facebook. Or if that doesn’t work, email”)
• Argument from ignorance (i.e., “If we don’t know it’s false, it must be true”)
• Reliance on anecdotal evidence, or the over-attaching of credence to heretofore unsubstantiated claims
• Being Susan Greenfield

I think the Baroness scores a perfect 10 on this one.

A BuzzFeed listicle in the making perhaps?

Brian Hughes

Brian Hughes is an academic psychologist and university professor in Galway, Ireland, specialising in stress, health, and the application of psychology to social issues. He writes widely on the psychology of empiricism and of empirically disputable claims, especially as they pertain to science, health, medicine, and politics.