Unlike people, who are constantly confused by their own biases and perspectives, numbers behave....if you know how to use them.
This is what I do for work every day:
First, I get what management thinks is the problem. Second, I talk to the people involved and find out what they think is the problem. Third, I get to retreat in to the numbers. I spend time looking at what we're doing, where we are, and where we'd need to be for everyone to be happy. It's the third part that's my favorite. No one argues, no political pressure, just puzzles, problems, and unexpected truths.
I use data every day to help improve health care, and I've been pretty successful at it so far. As I look around though, I realize how few people really understand the importance of good data in our lives. One needs look no further than election year politics to see bad data, poor interpretations of good data, and blatant misuses that make me cringe. In the healthcare realm, we don't have this luxury. I come from a world where you can't take chances, where misrepresenting your stats can result in very real human suffering.
This is why improper uses of data drive me nuts. Once you know what to look for, it's hard to stop seeing it. It's everywhere. Thus, I am giving myself an ambitious goal. It's no longer enough for me to use good data science for my own purposes. I want to educate others, and hone my own skills along the way. I want people to know what research is, how to read it, and how to question it. I want others to be as passionate as I am, and I want a place to vent about the reporting that annoys me.
Stay tuned.
My hero, Nassim Nicholas Taleb (who wrote "Fooled By Randomness" before "The Black Swam") noted that the military research people were the most open to criticism, because they got it that lives were on the line and you don't want to get things wrong. Similar to your "we don't have that luxury." Intellectuals - as a professional class, not as a compliment - can often go on forever theorising in a vacuum, about what should work, or might work, or darn it, just-would-be-so-cool-if-it-were-true. Which is why i hate them.
ReplyDeleteIn addition to your stats professor's three choices, there is also "Only apparently correlated." I tend to get fooled by that most often, leaping to conclusions. My freshman physics lab partner, who knew that all the correlations in the first half of the semester were going to be linear, would whoop "It's linear" whenever we got our second data point. "Why waste time?" he would ask.
On intellectuals: my brother Tim and I got a chance to talk quite a bit about my work at Christmas, and he asked quite a few questions. After I'd gone on for a while, I asked him for his thoughts. He just looked at me and said "It's really disturbing to me how differently you approach this from all the people in Washington I hear talking about it".
DeleteThings look different from the trenches.
Also, I love that last point. I'm stealing that point for a future post.
Was pointed at your site by maxedoutmama, looks interesting. I was raised an economist but understood that 6.2 billion data points, uncorrelated, meant most economics was a shot in the dark - after getting the requisite degree, a shot in the dark was more likely to be accurate. Read your profile to my 17 yr daughter, she said it sounded like her...when she was little and the ice cream was brought out, she did a dance...she still loves ice cream.
ReplyDeleteIn my experience, the ice cream/ice cream dance correlation at our house is at least a 0.83.
Delete...improper uses of data drive me nuts.
ReplyDeleteMe, too.
The worst offender is the economics profession, whose practitioners willingly torture some irrelevant data into proof of the desirability of proposed gov. programs/actions.
If economists wish to regain their integrity, they can start dropping the pretense that tests of statistical significance are the definitive analytical too.
Secondly, they can admit that all economists know with certainty is that competition benefits consumers, and that the economics professions has been complicit in countless interventions by Congress to prevent competition.