Closing Gap?

The gist of this article is that after years of measuring NYC's efforts to close the achievement gap showed consistent success, the gains made by black and Hispanic students relative to white and Asian students largely disappear when a widely accepted national measure is used instead of the state tests that had provided the earlier data. This is the kind of data that will be trumpeted about and said to mean many things, but I need to quote my statistics professor once more: "figures lie, and liars figure."

I do not just mean that some data is false or manipulated - that is clearly the case, as in the miraculous data that Houston produced in the 1990's that won, and eventually cost, Rod Paige his job as George W. Bush's first Education czar. Dropout/attendance data, mentioned in an earlier entry, are another notorious "depends on who's asking" kind of calculation. In cases such as these, where data translates directly into dollars (indirectly in "czar" jobs), the incentive to gather and/or report the data in a way that tells the story you want to tell seems hard to resist.

But there is another way in which these figures lie, and that is when numbers are used to give the veneer of truth to what in fact are points of view. For instance, charter schools are good. Charter schools are bad. Both points of view are often backed up with numbers of some sort. But the numbers - those that are available - do not support either conclusion. The only reasonable conclusion they support - so far, as least - is that some charter schools are good, some are bad, and some are neither.

So this paragraph in the NYT article stood out for me. "By 2009, the passing rates of black students on English exams had narrowed to within 22 percentage points of white students’, and within 17 points on the math exams. And charter schools, which predominantly serve black students, were doing so well that one Stanford University researcher proclaimed that they had practically eliminated the “Harlem-Scarsdale” gap in math."

I would argue that there are a few truths here - none of which can be accurately captured in a number-based argument.

1. Any measurement of learning measures what you measure, and does not measure what you don't. So it is no surprise that NYC kids did better on the tests their teachers were teaching to. Were those gains real? Probably, as were the gains of kids in other states with similar approaches. Would they have been just as real had they been teaching to a national standard? Probably...one compelling reason to adopt a national standard is that it enables apples to apples comparisons. Had we adopted such standards 20 years ago, we might have data now that gives us better insight into the real variables. Instead we have data that, in the sound bite form it will be used by many people, will cause folks to conclude that the achievement gap cannot be lessened, or that New York's efforts to do so are ineffective. Those assertions are incorrect - see below. (Such data might also enable us once and for all to accept that the achievement gap is mostly not about race, but about income...but let's not get too hopeful on that front.)

2. The achievement gap CAN be lessened, and as the data from the Stanford study clearly indicate, NYCs charter schools have done so more successfully and consistently than non-charter schools. This does not mean that charter schools "work." The data is the data: some of NYC's charter schools have been more successful in closing the achievement gap than most non-charter schools. We can then ask, "what about those schools causes that?" We actually know most of those answers: consistency of approach (the 3 Rs of rigor, relevance, and relationships); effective leadership and high morale, and deliberate parent engagement and accountability. We can also ask why some charters are more able to create those conditions than other schools (charter or not). Once we figure that out - OK, truth is we know most of that already, too - we could then go about addressing those underlying causes. So why don't we?

3. Because the debate in most places - at the levels policy gets made and carried out - is not about efficacy or meeting the needs of kids. It is about point of view: right/left, union/non-union, Tea Party/Green Party. It is also about job security: if your job is one of the ones that is safe only if the system stays largely as it is, how hard are you going to fight for something that changes that? While there will always be statistical trends that give us a sense of the right direction, there will always also be manipulation of data to fit the interests of whoever is citing it.

Here is a big-picture premise: reducing the achievement gap in a meaningful, lasting way is NOT going to happen if we do not reduce the income gap, the nutrition gap, and the widening community asset gap that concentrates more and more of those who've "made it" with folks like themselves, and pushes more and more folks who struggle to make it away from those places. We have decades - centuries, I'd wager - of data, none of which shows educational achievement moving in a way that is not correlated with those factors...at least not for very long.

But here is a smaller-picture premise: we do have data that says, both as a state and as individual communities, how we can invest our dollars, organize our schools, and build community and business partnerships in ways that produces better outcomes for more kids. Why we're not doing that does not have to do with how we interpret the data - it has to do with how willing we are to ignore what it tells us. As another professor once said, "the only problem with relying on common sense to solve problems is how uncommon it is."