EDUC 504/649: Accountability - An Alternative to Multiple Choice?

I keep coming back to (harping on?) this theme in various classes:

1. Yes, standardized tests can suck as a metric for real, well-rounded learning (e.g. can you write your way out of a paper back, do a high level math proof, or do actual science?)
2. Yes, there are reasons we use them -- parents and our communities deserve some way to see how we're doing with their children and their tax dollars (i.e. accountability) and standardized tests provide objectivity and efficiency in attempting to meet that goal.
3. What's the alternative that provides accountability, objectivity, and efficiency?

Courtesy of a post by the Learning is Messy blog, I found myself reading a Washington Post blog about a potential viable alternative, which led me to a Phi Delta Kappan article describing the alternative.

The article, entitled "Keeping accountability systems accountable" is by Martha Foote, and was published in January 2007. It profiles the New York Performance Standards Consortium, which is a coalition of 28 high schools in New York that focus on "inquiry-based methods of learning with classrooms steeped in discusion, project-based assignments, and student choice." These schools have have replaced standardized tests (such as the New York Regents Exam) with performance-bassed assessments, including

• Literary essays that demonstrate analytic thinking
• Problem-solving in mathematics that demonstrates high level conceptual knowledge
• Original science experiments that demonstrate understanding of the scientific method
• Social studiesl research papers that demonstrate use of evidence and argument.

The quality of the assessments is measured using rubrics. There are external evaluators involved -- experts in the disciplines and teachers from other schools. Overall, the system is held accountable by the Performance Assessment Review Board, which has lots of nationally known educators, business leaders, and public figures on it.

That all sounds great - I love the assessments! However, I also have some qualms / questions:

• The website is a little fuzzy on whether the "external experts" evaluate every student or only to spot check internal grading by teachers within the school. I think that for objectivity and cross-school comparability, probably external evaluators would need to be used all the time, but maybe I'm too cynical?
• It also doesn't really speak to how teacher feedback would feed into the projects (or not) -- does my teacher read a draft of my analytical essay or help me if I get stuck in my higher-level math proof? If they didn't help me make sure my science experiment was safe, they would be negligent. If my proposed procedure isn't safe in the first draft, do I fail, or am I helped? If I'm helped, is it still an authentic assessment of my learning?
• While 28 schools is a great start, it doesn't really speak to whether it would be realistic to roll this out over a whole state or (eventually) the whole country. I'd like to think the answer was yes, but I don't really understand it well enough to see how that could work.

Brian Crosby at Learning is Messy ties this and a recent article about standardized test scores declining in DC into a conversation about whether Race to the Top is encouraging real innovation or just helping schools making students better standardized takers. (As a side note: Second round RTTP finalists were announced today, and Michigan isn't on the list.)

What do you think? Could something like the New York Performance Standards work more broadly? Would it encourage the right kinds of teaching?