Friday, August 6, 2010

Standardized tests, school structures, and funding

Warning: thinking below is messy and evolving. But writing is helping me work my way through it, and any comments you leave can also be part of that process!

From the introduction to Teach Like A Champion by Doug Lemov (bolding mine):

So let us assume that students need to have both kinds of skills. They need to be able to read and discuss Shakespeare, but they also need to be able to read a passage they've never seen before and effectively make sense of its meaning, structure, and craft. They need to be able to write a short paragraph giving evidence to support a conclusion. They need to be able to solve for x. Most state tests do an effective job of measuring these skills, and while students who can demonstrate them are not yet fully prepared for college, there are no students who are prepared for college who cannot demonstrate them.

It's also worth noting that teachers who are better at teaching the skills measured on state tests are most often also the teachers who are effective at teaching higher order skills. I know this because within Uncommon Schools, when we correlate the success of our students on tougher internal assessments (essay writing assessments that are far more demanding that state tests, for example), there is a strong correlation between both the teachers and students whose results show the most growth and achievement on the two types. Furthermore, our teachers who achieve the strongest results from state assessments also have the strongest results in ensuring our students' entry into and success in college. In short, student success as measured by state assessments is predictive of their success not just in getting into college but of their succeeding there.

Finally, the correlation between success on even more straightforward assessments (nationally normed test scores) and ultimate academic success should be instructive to us. I often meet educators who take it as an article of faith that basic skills work in tension with higher-order thinking. That is, when you teach students to, say, memorize their multiplication tables, you are not only failing to foster more abstract and deeper knowledge but are interfering with it. This is illogical and, interestingly, one of the tenets of American education not shared by most of the educational systems of Asia, especially those that are the highest-performing public school systems in the world. Those nations are more likely to see that foundational skills like memorizing multiplication tables enable higher-order thinking and deeper insight because they free students from having to use up their cognitive processing capacity in more basic calculations. To have the insight to observe that a more abstract principle is at work in a problem or that there is another way to solve it, you cannot be concentrating on the computation. That part has to happen with automaticity so that as much of your processing capacity as possible can remain free to reflect on what you're doing. The more proficient you are at "lower-order" skills, the more proficient you can become at higher order skills.

So, to sum up, Lemov proposes that state standardized test scores

  1. measure knowledge and skills that are necessary but not sufficient for ongoing academic success,
  2. are predictive of success with higher order skills and higher education success
  3. measure foundational skills that need to be mastered in order to enable higher order thinking and skills.

This makes intuitive sense to me, although I would of course love to see the data supporting it.

I find this excerpt interesting because it articulates why I feel frustrated every time we start discussing in class whether using standardized tests to measure the effectiveness of schools is valid. On the one hand, no, standardized tests clearly cannot measure all the higher order thinking skills that we wish to develop in our students. On the other hand, if a school's students cannot pass the standardized tests, there is a problem. So while having 100% of your students score proficient on a standardized test doesn't necessarily mean that a school is doing a great job, having only 20% or 50% of your students score proficient on a standardized test does, in my mind, mean that the students are not acquiring the foundational competencies they will need in life, and are unlikely to be developing amazing higher order thinking skills while failing to be able to pass the standardized test.*

The question then becomes, can you blame the school for this? I think there's evidence that the traditional public school structure (180 days a year, 7 hours/day) often does an adequate job of educating middle and upper-middle class kids who mostly have a reasonably stable home life and families with enough money to not be worried about food, clothing, housing, and basic security (e.g. my own public school education in Bloomfield Hills). I think there's also solid evidence that that same structure is often inadequate in the face of the poverty and social disadvantages associated with the population in many urban school districts (e.g. Detroit).

So what happens when you have an inadequate structure and you impose high stakes standardized testing upon it? One of two things: you change the structure to something that works better, or you try to game the testing system. Charter schools like KIPP and Uncommon Schools change the structure -- longer school days, longer school years, highly disciplined approaches to teaching and learning -- and see good test results, which the author above ties to improved higher order thinking. There are of course other, more progressive approaches to changing the structure, such as adopting place-based curricula or addressing health issues and incorporating the broader community like the Harlem Children's Zone. However, the schools that are stuck with rigid curricula and inadequate instruction time and resources for their students to overcome the disadvantages the kids bring to school have no recourse but to try to game the system. This results in the much-decried teaching to the test, or focusing all extra resources on the "bubble" kids who will make the difference between meeting the average yearly improvement targets or not, or sometimes even cheating by teachers and administrators.

So if the end goal is to make the public education system serve all students at a level that at least lets them meet the minimum standards enshrined in standardized testing (and based on the arguments Lemov makes, I actually think that's a reasonable goal), the key question becomes how can we change the structure of the schools that aren't currently achieving that minimum standard. One approach is the extended school day and year and highly intensive teacher development and involvement embodied in the most successful charters. This is the approach the Obama administration seems to have latched onto, but it seems to have significant barriers to broad implementation in a public education system dominated by unions that would demand to have their teachers compensated for the extra teaching time in an economic climate where raising funding levels would be very difficult.

There are clearly other innovative approaches out there -- inquiry-based learning, place-based schools, and the Harlem Children's Zone. However, the advocates of those models don't seem to be making the case for scaling up and rolling out those models in other contexts, in a language that resonates with those awarding large amounts of private or public funding. I'm going to speculate that part of the barrier here is that the advocates of these models tend to be of a more progressive mindset, and tend to dismiss standardized tests. If instead the advocates showed that their approach of teaching beyond the tests also prepared students to meet the "table stakes" of proficiency in standardized test, they would have a much clearer way to "demonstrate success" and attract both the private funding that has been propelling organizations such as KIPP and the public funding currently being awarded through the Race to the Top and the Investing in Innovation (i3) funds.

What do you think?


* A 90% proficient rate might be okay, on the theory that 10% of your students might have test-specific disabilities that make it difficult for them to demonstrate their knowledge and skills in the standardized testing environment.

No comments:

Post a Comment