Friday, March 01, 2013

The limits of data and statistics

Recently, I ran across a couple of insightful essays by New York Times columnist, David Brooks.  Both dealt with the limitations of data.

Read his first piece here.

Read Brooks' second article here.

Interesting stuff.

I passed both columns along to a group of my friends.  Among the responses that I received is this one from one of my partners at CitySquare, Rev. Gerald Britt:

Brook’s column gets at something I’m trying to put some language around. You might remember an editorial in The Dallas Morning News that suggested that out of 3000 kids in 10 selected zip codes, only 26 graduated college ready. I asked a friend of mine, a former school teacher to help me with some research on these stats and she showed figures that revealed that in those same 10 zip codes, nearly 500 kids successfully completed their first year of college.

But she also said some other things in her reply that supports what Brooks said, “Traditional college readiness indicators for low income, minority, and ESL students do not measure motivation, persistence, and the ability of students to increase their ability to do college work over time. Research has shown for decades that SAT scores are not accurate indicators of college readiness for low income kids.”

“The college readiness standards used by the state are overkill. Admittedly, these high schools have a long way to go in recruiting teacher talent and programming, but the data used by StudentsFirst does not trace the successful kids in magnets and early colleges back to these zip codes.”

We here more and more about solutions to problems in education being ‘data-driven’, but, as Brooks points out – data is a tool and we’ve got to be attuned to more indicators than simply outputs and outcomes.


Anonymous said...

Strangely, perhaps, I think statistics and anger have something in common. As a song says: "anger can be power if you know that you can use it." But if your anger is controlling you instead of you controlling it, you're sunk.

Similarly, if you are using your statistics, that's great. But if your statistics wind up calling the shots, its all downhill from there.


Cathy Sweeney said...

Mr. Brooks and Rev. Britt make insightful points, and StudentsFirst has a viable and necessary cause to support. The unfortunate reality is this: the bigger the population, the more likely a decision maker is going to manage to the bell curve.
What's needed to look 'outside the bell curve' is consideration for the intangible assets that Rev. Britt and StudentsFrist support: resiliency, motivation and 'sticktuitiveness.' What's needed is for decision makers and their staff to build relationships with individuals, not with zip codes.