"A child's learning is the function more of the characteristics of his classmates than those of the teacher." James Coleman, 1972

Tuesday, July 03, 2012

Guest Post: The Education Sector Attempts to Head Off a Common Core Trainwreck

Dr. John Thompson was an award-winning historian, lobbyist, and guerilla-gardener who became an award-winning inner city teacher after crack and gangs hit his neighborhood.By John Thompson

The Education Sector's new report, Getting to 2014 (and Beyond): The Choices and the Challenges Ahead, could have been entitled "Whistling Past the Graveyard." The report starts with the standard proclamation that "new accountability systems will reflect the expectation that tremendous progress in students outcomes across the country could occur through the unprecedented efforts unleashed by Common Core." Yeah, and perhaps a train wreck could be avoided if accountability hawks repudiate their tactics of the last generation, while dealing with the nine issues described by the Ed Sector. So, let's take a look at what would have to happen to avoid one disaster when value-added meets Common Core.

I have been pestering policy types for a scenario where value-added evaluations could co-exist with the transition to Common Core Assessments. I have usually been met with a sputter, sentence fragments, and an admission that we will have to choose between one policy or another. Rarely do these reformers get past the most obvious roadblock. Many or most states will soon replicate the experience of Tennessee where proficiency on eighth-grade math fell from 90% to 26% in two years when standards were raised. Needless to say, the panic and the retribution that these drops will prompt will not be conducive to rational decision-making.

The transition from bubble-in accountability to its opposite, the assessment of college readiness skills, will require teachers and principals to buy into a mixed message. During the next two years, educators will receive training for what Craig Jerald calls a "massively difficult, change in instructional practices." In the meantime, they will have to teach to, and be subject to termination by, primitive standardized tests of basic skills. That might be hypothetically possible if, to borrow from Michael Cohen, accountability is transformed into "something meaningful," if everyone recognizes that Common Core could produce "tremendous progress in student outcomes," and if we all rally toward this vision.

To his credit, Bill Tucker is, to my knowledge, the first "reformer" to articulate a scenario where value-added evaluations do not need to be abandoned in order to implement Common Core. Tucker maintains a brave front, but I suspect that he, like Jerald, does not really believe that such a two front war is possible.

Tucker cites testing expert Drew Gitomer who explains that we can't assume that the "patterns of growth from one test to a dramatically different test the next year" will be comparable. "If the new standards and tests meet their expectations and really are measuring different learning outcomes," Tucker notes, "'then a whole lot more needs to be understood before we simply apply value-added models to these data.'" The most likely outcome, I would add, is that it will prove much harder to raise Common Core scores in high-poverty schools.

Tucker then crafts a metaphor that nails the dilemma:

If today's value-added measures are like calculating a sprinter's improvement in the 100-meter dash across different tracks, then in spring 2015, we'll be attempting the equivalent of calculating value-added across not only different tracks, but also racing events. And while we may be able to infer a number of things about a 100-meter sprinter's improvement by her performance the following year in the halfmile, calculating a precise measure will be extremely difficult, at best.

If we accept Tucker's professed assumption that data-driven evaluation systems must remain in effect, then he presents the only plausible roadmap. Reformers would have to make a "bold statement" about the importance of Common Core and announce a "hiatus" in test-driven evaluations. I can't believe that Tucker really thinks it would be possible to stop value-added evaluation for only one year, but he keeps a brave face in calling for that brief lapse in test-driven accountability.

For the hiatus to remain a hiatus, and not the abandoning of value-added evaluations for multiple year, or forever, all the stars would have to align perfectly. Fields trials would have to remain on schedule and produce reliable results. Statistical models that have been developed over two decades for one type of test would have to prove reliable for a completely different test, as risky innovations are fast tracked. Even when using seven years of data, and using the latest in a series of experimental models, the Los Angeles VAM was only 27% accurate in placing teachers' performance in the correct evaluation category. But, educators under pressure to completely transform the way that they do their jobs, especially in the poorest schools, would then have to submit to high stakes evaluations using one year of real-world data.

Value-added models (VAMs) using old-fashioned test scores have yet to control for concentrations of special education students, English Language Learners, and other peer effects, or for large percentages of top students who already score highly. I am most concerned that value-added will produce an exodus of the top teachers from low-performing schools where it is more difficult to raise test scores. For that reason, Tucker's article should be read in context with the other great piece, by Robert Balfanz. Balfanz asks how the 15% of high schools with the greatest problems with chronic absenteeism and the highest dropout rates could continue to improve graduation rates while biting off the completely different challenge of Common Core. Balfanz writes, "if we continue to concentrate the neediest students in a subset of schools that are not designed for success, we will neither be able to raise performance levels nor graduation rates for these students." Balfanz then recommends three more promising approaches, using completely different strategies, to help the poorest schools.

But Balfanz's astute recommendations are the topic for another post. The point in regard to Getting to 2014 (and Beyond) is that educators already have far too much on their plates, and that is even more true in high-poverty schools. In a rational world, value-added evaluations would be the first thing we would take off teachers' and principals' plates.

Tucker and the rest of the Education Sector contributors focus on "a new expectation of accountability systems" with the assumption that they are the only way "to continuously improve their ability to drive students toward college and career readiness." If we really want to improve our poorest schools, we need to listen to Balfanz and others while not assuming that their research-based policies can only be implemented when shoe-horned into the failed ideology of "accountability."

Dr. John Thompson was an award-winning historian, lobbyist, and guerilla-gardener who became an award-winning inner city teacher after crack and gangs hit his neighborhood. He blogs at thisweekineducation.com, and huffingtonpost.com, and is writing a book on 18 years of idealistic politics in the classroom and realistic politics outside.

1 comment:

  1. Why can't they just let the students think beyond the railed principle of test-based education? It's killing the students' critical thinking abilities and creativity, and that is what's eating off the quality of our graduates.