David Brooks, Data-ism, and Why You Shouldn’t Listen To Empty Suits

…including this one.

It seems that I and my fellow nerdlings have struck a bit of a chord over at the NYT–such a chord that David Brooks just had to respond with still more evidence that he does understand data. He does. He does. He does.

No. He doesn’t.

Sunlight is still the best disinfectant, but it’s better to expose methods rather than facts. Facts, as Brooks so elegantly points out, are open to interpretation. It is more robust (and more honest) to form a hypothesis then look for proof against. A hypothesis has to have a yes/no answer. I’m sure that’s terrifying for someone who’s used to dealing in “maybe”(s) and “might”(s) and “let’s just consider”(s).

So, let’s try it on Mr. Books.

David Brooks Hypothesis: “Your brain is bad at math.”

False: Actually, your brain is quite good at math, when it’s allowed to process numbers intuitively rather than in the rigid manner taught in grammar schools. In one experiment, a native tribe with no number words past five did as well as we do with approximate numbers and better at reasoning with logarithmic scales. This kind of math reasoning, one could argue, is much more important for our everyday decision-making than being able to calculate the square root of 437.

David Brooks Hypothesis: “Data struggles with context… People are really good at telling stories that weave together multiple causes and multiple contexts.”

False: Narrative Fallacy. QED.

David Brooks Hypothesis: “Data creates bigger haystacks…Falsity grows exponentially the more data we collect.”

False: The problem is misused methods, not the amount of data. There are many, many statistical methods that account for multiple hypothesis tests. In fact, the problem to which Brooks alludes is more about our need to see patterns where none exist, and that has nothing to do with how much or how little data we have.

David Brooks Hypothesis: “Big data has trouble with big problems. …For example, we’ve had huge debates over the best economic stimulus, with mountains of data, and as far as I know not a single major player in this debate has been persuaded by data to switch sides.”

False: The IMF has switched sides. The people who haven’t switched sides have not failed to do so just because they lacked data. Already handled by someone with better data than mine.

David Brooks Hypothesis: “Data favors memes over masterpieces.”

False: Infographics are being used to explore literature and illuminate what makes  some pieces so moving. Shallow data analysis favors memes over masterpieces, just as shallow reading does. 

David Brooks Hypothesis: “Data obscures values.”

Data can obscure values. Data can also illuminate them.

David Brooks Hypothesis: People like Melinda think data is the only tool.

False: Melinda believes in art. Melinda believes in beautiful prose (and does not want to live in an overly data-driven world where people allow themselves to be herded by statistical models). Melinda believes in music (though she can’t make any of her own). Melinda understands that it is difficult to make decisions, even with all the data in the world.

Melinda does not, however, believe that David Brooks has anything interesting to say on the subject of data. Hypothesis tested and confirmed.

To the regular readers of this blog: Look, this is semi-personal. I think David Brooks is a stuffed shirt and a windbag, and then he had to go bringing that weak stuff into my house because Nate Silver was too successful to ignore, and I could not let that go. (Because I’m like that. That’s why.)

However, I’m betting there’s a David Brooks in your company, in your family, or in your state legislature. I bet there is a HIPP stomping all over you with his HIPPO right now. I bet he/she is telling you you’re wrong when you’re right, grabbing your point and beating you with it, and generally sucking all the air out of every room you enter because that is the only way he/she can win.

It does not have to be that way. Grab your slingshot. Follow me or forge your own path (I prefer the latter), but stop letting empty suits with nice ties tell you what to think. This is a unique moment when there is suddenly a whole lot of interest in  measuring outcomes and testing processes. Do not waste it.

Don’t let the HIPPOs bully you back into your basement. Gently, politely, graciously offer to test any hypothesis that comes your way. Run the test, admit when you’re wrong, but do not back down when you’re right. If your organization doesn’t want to test, start looking for a better job.

There are so few opportunities in this life to make the world a better place. Don’t let this one pass you by.

About Melinda Thielbar

Melinda Thielbar is a co-founder of Research Triangle Analysts, Ph.D. statistician, spinner of fine yarn, martial artist, fraud analyst, and fiction writer. In other words, she's a polymath. Follow Melinda on Twitter @mthielbar, or join the Research Triangle Analysts group on G+ to join the conversation about data science.
This entry was posted in Uncategorized and tagged , . Bookmark the permalink.

7 Responses to David Brooks, Data-ism, and Why You Shouldn’t Listen To Empty Suits

  1. Cat says:

    Someone sent me this article earlier today and I was just shaking my head (and wondering if you were reading it). Due to his ignorance, he makes some misstatements (for example, confusing type-1 error with small effect sizes; assuming nobody performs honest assessment; treating a designed experiment, applied to “the economy,” as the only way to learn something about “the economy”). But he sounds very authoritative, and so someone who can’t critically evaluate his arguiments is likely to believe what he says. This will not likely be corrected in the NYT.

  2. Great job Melinda, I really appreciate your response to Brooks! I agree with Carla, you should write a letter to the editor.

    As David was putting one foot in his mouth, he was shooting himself in the other. His attempt to pit humans against data and declare the former victor only bolsters the argument in favor of data science. People add value to data, not the other way around, which is precisely what is stressed by practitioners of the field. Data science does not aim to replace people with data, but rather decisions that don’t consider the data with decisions that do.

  3. Phil Simon says:

    Whether he’s right or not I don’t know, I just think that when the NY Times discusses data and Nate Silver drives much of its traffic, good things will result.
    Data–Big Data in particular–is becoming too big to ignore.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s