Everyone seems fired-up about multiple measures etc…Sherman Dorn has a post that is well worth your time, and I think his conclusion is a pretty good educated guess…just sadly 7 years too late, but better late than never.
Meanwhile, at the HuffPo, Jerry Bracey is wallowing in pity. But, cutting through it he seems to be coming out against multiple measures, but more for capacity reasons. Bracey knows this stuff so I’d be interested to know what sort of multiple measures system he thinks could stand up, near term capacity issues aside? Outside of things like AP and IB, I’m pretty skeptical in the accountability context and would hate to see the hard won progress of 1989 (Clinton and Bush I), 1994 (Clinton), and 2001 (Bush II) on common standards for all kids eroded.
About all of this, three things jump out at me. First, as my colleague Tom Toch has reported, the testing industry is stretched. But, what I don’t know is on a per capita basis how many more (or less) errors are we seeing? Is the industry getting worse or have they always had these mishaps. I do recall problems prior to 2001, so what’s the trend? As a state policymaker I can’t even get a decent sense of what companies are better or worse on say, errors per 100,000 standardized assessments given. Who are the worst offenders and best corporate citizens here? And what states have the most effective enforcement mechanisms in place to hold companies accountable? There are penalty clauses but considering the scale of these contracts they just might be considered costs of doing business. And, when there is an error everyone jumps up and down but sometimes in terms of the overall percentage of tests given, it’s pretty minor, what are those numbers? That’ll either be headline or snooze inducing depending on the data…
Second, a lot of stuff gets conflated in this debate and blamed on No Child Left Behind. For starters, on this curricular narrowing point, make no mistake, pre-2001 (when No Child was passed) was hardly a golden-age for high-poverty schools where kids were getting a rich education complete with arts, music, well rounded physical education, and a great college-prep curriculum. NCLB does get hit with a lot of correlation-causation attacks. And, narrowing at the expense of content backfires anyway. In addition, while I don’t want to see all multiple choice tests, an ethic is springing up that multiple choice is all bad. That’s actually not the case and you can have high quality multiple choice tests. E.D. Hirsch has written about this.
Finally, this is an archaic technology, no? Surely there are better ways to assess kids than pencil and paper tests that can get information back to teachers faster, in a more useful way, with less time spent on testing, and so forth. I look at what, for instance Wireless Generation or Grow, are doing and wonder why we’re not looking further ahead? But there is not a lot attention being paid to all that via public policy. Instead, we’re really creating incentives to keep the world as it is and I worry that we could be on the verge of trying to invest a great deal to make the very best gas lamp the world has ever seen when electricity might be just around the corner…In a way it’s like the aerospace industry in the early 1970s when the competition was all about making planes that were a little faster or lighter through marginal advances in metallurgy or fuel mixes. Then along came Skunk Works with a plane that no one could see — stealth technology. What’s our “stealth-like” breakthrough on this issue and how do we support the work to get there?