Wednesday, April 2, 2008

Lies, damn lies, and statistics

As a follow up to yesterday's story about low graduation rates in city school districts, there's this from Columbus Public Schools ( via 10tv ):

According to the group's findings, Columbus City Schools graduated 40 percent of high school seniors in the 2003-2004 school year, Landers reported.

However, administrators with Columbus City Schools said the graduation rate that year was about 60 percent - slightly less than the national average.


So what accounts for a 20% discrepancy in droupout rates? From the same article:


States calculate their graduation rates using all sorts of methods, many of which critics say are based on unreliable information about school dropouts. Under No Child Left Behind, states may use their own methods of calculating graduation rates and set their own goals for improving them.

Columbus City Schools Associate Superintendent Mitchell Chester said the state would soon adopt a new way to calculate its graduation rates.


Oh gee. So the schools get to make up how they report statistics? Wow.

It's a truism that if you can't measure something that something can not be measured. If one gets to redefine how things are measured, then management isn't possible.

The point of management, in this case, is to improve schools based on measurable outcomes. If school districts don't want to "look bad", they should improve their performance as best they can not fiddle the numbers.

Looks like some in the government agree.

I've no idea if that will work, but it's a step in the right direction. At least, if we're to be saddled with the tax burden of public schools we, as a society, should get some return on our investment.

Schools should be more than a holding pen; of course, with drop out rates so high one could argue they're not even good at that.

No comments: