Jun 09 2013

If your “science” has no data, no one should believe you

The British writer Martin Robbins has a long and important battle against what he terms “data-free celebrity science”. This is “science” as touted by those who have established themselves as good scientists or thinkers – among their colleagues – and proceeded to use that goodwill to build a soapbox to spread their ideas.

In his new post on the latest offender, Oliver James, Robbins writes:

[This approach to science] has some distinct advantages for the academic, as Oliver James admitted in response to an impertinent audience member who challenged him at the Hay Festival. “I myself am a busy man,” The Telegraph reported him as saying, “I don’t have time to muck about doing intervention studies.”

Interesting that he describes doing science – gathering data, evidence, testing and verifying claims – as “mucking about”.

Assuming you’re not sucking thumbs, these ideas are worth examining via peer-review and proper journals. In 2011, this was reiterated by the excellent Ben Goldacre, author of Bad Science (the book and blog),  when he called out the most prominent offender of data-free, celebrity science: Baroness Susan Greenfield.

Greenfield, who is a renowned and certified scientist, has for some time been making wild claims regarding the relationship between things like technology, the impact on the brain, attention and knowledge-retention, etc. Computer games cause dementia! Internet something, something, autism!

Goldacre wrote:

Why, in over 5 years of appearing in the media raising these grave worries, has Professor Greenfield of Oxford University never simply published the claims in an academic paper?

A scientist with enduring concerns about a serious widespread risk would normally set out their concerns clearly, to other scientists, in a scientific paper, and for one simple reason. Science has authority, not because of white coats, or titles, but because of precision and transparency: you explain your theory, set out your evidence, and reference the studies that support your case. Other scientists can then read it, see if you’ve fairly represented the evidence; and decide whether the methods of the papers you’ve cited really do produce results that meaningfully support your hypothesis.

Perhaps there are gaps in our knowledge? Great…a scientific paper is the place to clearly describe the gaps in our knowledge, and specify new experiments that might resolve these uncertainties.

After all, these are clearly public health issues. Either they exist and require proper identification, via peer-review and evidence gathering, or they’re not and thus we are wasting time on nonsense.

Greenfield offered a strange reply to a reporter who asked what her hypothesis really was regarding computers and the brain. Bizarrely using Richard Dawkins and his atheism writing, Greenfield said:

“How can I publish just one paper?…Does Professor Richard Dawkins have one paper which sets out why God doesn’t exist? Tell me one experiment I should do that proves once and for all whether computers are evil for the brain.”

One experiment?  “Once and for all”? Unless we’re talking deductive reasoning – “All swans are white” – then one experiment probably won’t do much and no scientist should think otherwise. Science frustrates people because it is tentative, because it is often counter-intuitive, because it requires others picking your hypotheses and theories apart.

Science, as far as I can conclude, is merely good reasoning writ large. If it’s true, experiments and evidence will demonstrate; if not, then they will do the same. This allows us to spend resources – like time, money and attention – appropriately. Do we spend it on vaccines that have proven to save children’s lives or on ceasing such operations because vaccines cause autism?

I see little difference between someone claiming they dance with unicorns on a chocolate moon and someone claiming Mario gives you dementia. Neither has evidence. Recently, when I asked friends where their evidence was to support their website’s claim that they were the Number 1 group in their field, they got defensive; others laughed and began playfully mocking me, calling me names like “Sherlock” and claiming my catchphrase is “WHERE’S THE EVIDENCE!”. Well, good. Let that be a catchphrase for us all. We should try to have evidence for our (empirical) views – otherwise on what basis are we believing them? It makes no sense to believe things which have no rational, (and if possible) evidence-based justification. Otherwise, we could simply “choose to believe” whatever we want. Of course, we can’t have evidence for “everything”, but there’s no reason why can’t strive to, encourage others to do the same and welcome rational criticisms to those beliefs.

Of course, all this only matters to those who care about believing what’s real rather than what’s comforting or adheres to your agenda. James, Greenfield, etc., know how science works but appear unconcerned with actually responding to the most fundamental aspect of science and what anchors such a worldview – answering: “Why should I believe what you say?”


Skip to comment form

  1. 1

    But… “where’s the evidence?” is my catchphrase. At least it seems like it some days.

  2. 2

    For me it’s usually “Sources!!1!” especially if I’m arguing on Facebook.

  3. 3

    Of course, some claims are unsupportable because the claim is essentially meaningless. It merely appears to be packed with an objective meaning.

    Like the claim to be “number one” in a certain field. What does “being number one” consist of, exactly? I’m “number one” in my field — which I can support by defining my field so specifically and narrowly that I essentially define myself into being “number one”. But that’s completely meaningless unless the person on the receiving end has agreed that my definition is the accurate one. Being “number one” is meaningless because it has no objective standard upon which to be based. Even though it appears to be so, it’s not a claim that can be objectively measured.

    It’s the same with claims made on TV for vitamins and other OTC products. In this case, the loaded word is “supports”. What does that mean, exactly? “Supports”. Turns out, it means nothing. Zero. Bupkis. It’s a claim that is true by virtue of having no meaning and no standard against which it can be judged. Which is why it’s the word the FDA lets these folks get away with in advertisements. Which is why any time you hear the word “supports” in any advertisement, you can translate it to mean “does nothing”. Because if it actually worked, it would be a supportable claim and the word “support” would be substituted for an actual claim.

    In the advertising biz, they call it “puffery”.

    Like the Papa Johns slogan “Better ingredients, better pizza”. Well — they were challenged as to whether their ingredients were really “better”. Turns out, their ingredients are to a close approximation exactly the same as every other pizza makers’ ingredients. And the word “better” isn’t meant as a comparative against other pizza makers, but against … well … against ingredients that aren’t used to make pizzas. Like cow dung. They want you to believe that it’s a comparative statement; but when challenged, they acknowledge it isn’t.

    TL:dr: Fucking language. How does it work?

  4. 4

    I agree. I’ve said the same thing about other concepts, like “the patriarchy” and “rape culture”, where I only get snarky replies in return but no data to back it up. If it’s a science, it should have sufficient data to support it, otherwise why should we take it seriously?

    As Hitchens once said, “What can be asserted without evidence, can be dismissed without evidence.”

  5. 5
    F [i'm not here, i'm gone]
    “How can I publish just one paper?

    Who limited her to one paper? Why would any scientist want to publish only one? Of course, where to publish such a paper might be problematic, as The Journal of Moral Panic Quarterly is not yet in print.


    I also agree with one of Pitchguest’s points, and therefore would like to see some evidence for the assertion that there is no evidence (or that such evidence isn’t conveniently packaged and delivered to Pitchguest) for patriarchy and rape culture. And possibly also evidence that the sky sometimes appears blue and that humans have actually visited the moon.

  6. 6

    F [nucular nyandrothol]:

    Followed by another snarky reply. Should I rest my case or keep going? Sadly no evidence have yet to be reported (for a concept so overbearing like “patriarchy” and “rape culture”, it should have peer-reviewed papers published in the hundreds) but then the burden of proof doesn’t fall on me. Which is funny, because in the OP the author states he asked for evidence and got playfully mocked for it. Which is why what you’re doing right now. Ironic.

  7. 7

    ***which is what you’re doing right now (minus the why)

  8. 8

    Sources and methods, nothing else. Some people know that their brand of pseudoscience is quackery and so they play games to get uninformed people to listen to them. Science is all about verification of methods and reliable sources for those of us who learn about it second hand. If it’s unverifiable it’s not science, the claims of atheism are not science. Biological evolution is science, neurology is science, the Big Band is science but no one has devised a suitable experiment to figure out if god exists, but Dawkins or any other atheist will tell you that. Dawkins informed us of that fact in his The God Delusion but I don’t think Greenfield has read that.

  9. 9

    Pitchguest is, of course, not looking for evidence of “patriarchy,” but rather evidence of a massive conspiracy by men to oppress women. He uses the fact that evidence for the latter is not forthcoming to dismiss the former, hoping all the while that the readers are not aware that “patriarchy” is not a synonym for “conspiracy.”

  10. 10
    Random Bystander

    A bit of switchery in Pitchguest’s original comment that should probably be noted:

    The OP referred to “hypotheses” and particular claims. Pitchguest referred to “concepts”. A concept is not a hypotheses. Hypotheses can be supported (or undermined) by evidence and experiments. Not all (or even most) concepts can, as many concepts are not claims.

    It is important to note the switch here. If Pitchguest had stated some particular claims that he couldn’t find support for, someone could, maybe, provide that support — but he didn’t. Impressively convenient, that.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite="" class=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>