Monday, August 29, 2011

Correlation vs. causation

This is what I wrote about correlation vs. causation in 2007:

Correlation vs. causation


Last week's front-page WSJ article about the controversy integrating schools in Milton, MA ( got me thinking about a joke I first heard earlier this year, which I found hilarious:


At a U2 concert in Dublin, Bono asked the  audience for a moment of silence.   When the room became perfectly quiet, he began to clap his hands slowly.  Then, into the microphone, he said, "Every time I clap my hands, a child in Africa dies."  After a few moments, a lone voice from the crowd pierced the profound silence: "Fookin' stop doing  it then!"


The reason this is so funny is that the guy confuses correllation and causation (another classic example is that flood deaths in Bangladesh are highly correlated with ice cream sales in NYC, but nobody thinks that our consumption of ice cream is killing Bangladeshis -- it's just that both peak in July and August).


While these are examples are silly ones, it's actually very common for people to confuse correlation and causation, which is, in part, what I think is going on in Milton, MA (and around the country): perfectly sensible people look at the fact that: 1) primarily black and Latino schools, on average, do very poorly academically and 2) primarily white or Asian schools, on average, do much better academically, and conclude (often subconsciously) that black and Latino students cause schools to be bad, and that therefore the best solution is to send as many black and Latino students as possible to primarily white/Asian schools. 


I am not making this up: here's an excerpt from Wendy Kopp's commencement speech at Mt. Holyoke last May (


Most Americans view educational inequity  as an intractable problem.  Every year, Gallup surveys the public, asking why we have low educational outcomes in low-income communities.  Out of twenty options, the public responds  'lack of student motivation,' 'lack of parental involvement,' and 'home-life  issues.'  In other words, most Americans believe this is an  entrenched societal problem rather than a problem that our schools can change.


In other words, Americans have it precisely backwards: they think schools are being victimized by lousy students, parents and communities, rather than the reverse!


To be fair, as I've noted many times before, it's really, really hard to operate a safe, effective public school in an inner-city community, but it can done.  It must be done!  As for those so-called unmotivated students with uninvolved parents and home-life issues, go visit any No Excuses charter school (like KIPP's Open House this Wed.) and you'll see that the right school can turn those same kids into hard-working, high-achieving, college-bound students with engaged parents.

 Subscribe in a reader