Bob Wise, the former governor of West Virginia, is now president of the Alliance for Excellent Education. In that post, he's had an epiphany of sorts about the importance of developing better data systems for government. States, counties and cities are accelerating the rate at which they germinate statistics, including kudzu-like standardized tests. But creating new measures is only half the battle. And the battle is pretty much lost if there's no way to use new information to measure educational progress. That, in turn, requires better ways to gather the material, analyze it over a period of time and disseminate it.
Wise's broad interest in education is no surprise. One of his most significant contributions as governor was to create the PROMISE scholarship, designed to make college education affordable and keep West Virginian students at home. But he concedes that his new fascination with gritty numbers and grids is a bit of a shock. Says he, "If I'd gone to the fortune teller three years ago and she said, 'Bob you're going to be passionate about longitudinal data systems,' I'd have said, 'I want my money back.'"
His work couldn't be more necessary. Even as simple a piece of information as dropout rates turns out to be mired in the muck of mathematical opacity. In fact, the states' dropout rates are far higher than was once believed and this critical situation "has been obscured" in the words of a 2006 Alliance for Excellent Education report, "by inaccurate data, misleading official graduation and dropout rates and flawed accountability systems at the state and federal level."
--Katherine Barrett and Richard Greene What have you learned about the need for better educational measurement systems?
When I was in political office, I wish I'd had good data. In the last two years, I've come to appreciate how critical data is. Developing good longitudinal data systems will have an impact from the classroom to the congressional meeting room. Teachers can improve their teaching practices to a particular student and legislators can make multibillion-dollar spending decisions based on good data.
In the legislature and the Capitol, data systems aren't necessarily the top priority. If you say, I have $3 million to develop an innovative program or data infrastructure, guess which one gets short shrift?
Data, in and of itself, doesn't sound very exciting. But as every state moves to a standard-based system, data is critical to any kind of decision.
What's driving the need?
This is an amazingly mobile society now, where a family is likely to move a number of times as the child moves through the school system. If they're moving to the next district or across the country, you want to make sure that data follows them, so that people can assess them and determine what they need for a quality education. If you're talking about a catastrophe, such as Katrina, you want to have data systems that reflect where that child is and data systems that can talk to each other and communicate.
Today, 80 percent of current jobs require post-secondary education, which means high school graduation. Also, one thing that states do which I admire is copy each other. I loved the NGA [National Governors Association] because it gives you the chance to steal the best ideas and take them home. But you want to make sure that the ideas you're implementing are backed up by sound data.
Fifty years ago, maybe it wasn't important. But today, what we recognize is that we need to make sure that every child in every school district has a certain level of quality education.
A couple of years ago, the nation's governors all agreed to work toward comparable graduation rates. Are we getting close?
We're a lot closer than we were, but we've got some ground to cover. The NGA compact sets the stage. Having said that, states have announced different implementation plans. Only one state has codified the governors' agreement and that's the state of Maryland.
[And while] codifying the governors' compact is an important statement, it only has the strength of that current governor behind it. It's not law. The governor can change or it can slip off the radar screen. For a whole lot of reasons, the federal government should say we recognize what the governors have done. It should be an element of No Child Left Behind. I believe it should be a stronger element of accountability than it currently is.
What's holding this effort back?
Dropout rates don't have the same weight as an accountability element in NCLB. They're getting less priority. There is reference to them in the law, but in practice they're not a major element. Each state is still able to determine its own graduation rate and if each state is determining its own graduation measurement, you can have wide variances.
We run our kids in a mile race and assess them rigorously every tenth of a mile and don't bother to see if they've reached the finish line. If you're not going to be counting how many finish, then that's an incentive to push people out who aren't testing well.
If you have a longitudinal system, you'll have a much more robust or more accurate calculation, but most states aren't there.
We know that a longitudinal system generally depends on using a student identifier so you can track the same student from 9th grade through the rest of high school. A lot of people are nervous about the privacy implications. Is there a way to develop useful longitudinal data, without that?
I don't see how you'd do this without a student identifier.
I think privacy issues are very significant. I'm sensitive to privacy issues. But the privacy of what? A student identifier is used to the student's advantage.
It can be used to track your educational progress, so that when you move, the new school district is aware of where you are.
It's also used so that when you move to the next grade or if your teacher changes, the new teacher can look back and see what you need. I understand that children aren't widgets, but I don't know any business that would not have a way of identifying its employees and their progress or the products created.
The student identifier is necessary to be able to deliver the best services to the individual wherever that individual is or goes. And, in a cumulative way, to determine what is working and not.
I can't even conceive of having a system that would say we're going to spend hundreds of thousands of dollars per student and move them through 13 years of school and then not measure how we did it.
What are the obstacles that are impeding progress on this issue?
Data is not a sexy issue. Dollars are an obstacle. You need training in setting this up and in administering it. It's an initial cost, but the return on investment we think is many fold. You avoid bad decisions and it helps you make good decisions and you can improve your real-time instruction. A teacher will be able to get immediate return on an assessment, rather than waiting until the child has moved on to the next year.
States have varying degrees of will. It's important to build a data system and one that's not just sitting in a basement. Everyone has to be trained -- from teachers to administrators. It's a question of committing effort and resources. Once you do, you'll get a great return. You can truly evaluate where you're doing well and where you'll have to improve.
Do you think poor data in the past has impacted public policy and management?
The substitute for data is anecdote. About 90 percent of legislation is already done by that. In the hours before the bill is worked out, someone says 'In my school I heard that this happened,' and that gets written into the law. That's why data is increasingly important. Because there are billions of dollars of spending decisions at stake, legislators and members of congress are increasingly demanding to see the results. They also understand that the rose-colored numbers that were coming out ten years ago about dropout rates simply aren't holding up.
Are you optimistic?
I think we're making progress. There's a lot of interest on Capitol Hill on how we adequately measure the finish line. It's not the graduation only, but the quality that's behind the diploma. If it doesn't reflect a rigorous curriculum, it doesn't mean that much.
In graduation rates, we're making progress and the administration has been supportive in assisting states in developing quality data systems. Twenty-seven states have received grants to develop longitudinal data systems. That's the area you can get the most bang for relatively few federal dollars. This will permit determinations of what's working and what's not and it will help make future decisions about allocations of resources. Of the 27 states that received grants, Florida is the farthest along. They are able to make decisions at every level on a much more informed basis.
It's nice hearing a former governor talk about the importance of good data since we've been talking about the importance of good performance measurement data for about twenty years.
I think your moment is coming. All the focus that the president is giving and the governors are giving on international competition -- it's a community that wants solid data.