Posts Tagged misusing test scores

A Punitive Mindset Ruins Good Testing Data

Published by — Standardized tests are fantastically useful.  But schools can’t be harangued, threatened or sanctioned into improvement.


The fault, dear Brutus, lies not in our standardized achievement testing, but in what we do with said testing.  Achievement data are fantastically useful.  These days virtually every industry collects and analyses the best information available to make smart decisions.  So the recent groundswell aimed at ending the flow of testing information is akin to insisting we all stick our fingers in our ears and holler:  We don’t want to hear it!

Easy now.  Let the data speak.  Just quit jumping to ill-considered actions.

But the situation has gotten so bad, parents are refusing to allow their children to take standardized achievement tests.  Congress, seemingly stuck in brute partisanship, is arming for war over the reauthorization of the Elementary and Secondary Education Act, known since 2001 as No Child Left Behind (NCLB).  One huge battle will be testing versus no testing.  Likely it will focus on money, although the $34 per kid for state and federal testing seems a small price to pay to find out if the taxpayers’ $600 trillion dollar annual investment in public education is getting results.  That’s trillion with a “t.”  We can spare $1.7 billion, with a “b,” to do a check-up.

How testing became a monster.

The one great thing NCLB did for the nation was goad states into building robust data systems.  Though far from perfect, even state-designed testing programs surfaced glaring racial and socio-economic disparities.  Tests revealed that special-needs children and non-native English speakers often languished in segregated programs, “protected” from higher expectations.  It wasn’t pretty.  Still isn’t.  But sunlight on the plight of the underserved inspired a lot of creative thinking about how to narrow the gaps.  As a nation, we’re grown uncomfortable with these disparities.  And that’s a good thing.

A very bad thing, though, was that punishment was baked into NCLB from the get-go.  In the name of “accountability,” the 2001 law disciplined failing schools with an a menu of escalating sanctions including humiliation and threats of state takeover — as if states had the capacity to take over schools.  Annually, all schools had to meet rising achievement benchmarks with goal of having all students “proficient” by 2014.  Each year that a school bombed its benchmarks, the feds and states had license to trumpet failures in the media, impose insulting oversight, and force the schools to write the parents about their failures.  NCLB was a big, bad Dad that believed he’d get results by yelling louder and getting meaner.

Everyone knew that achieving nationwide proficiency in 2014 was statistically impossible.  But oh well.  States issued new naming-and-shaming reports anyway.  Adding insult to injury, those most affected were low-income kids, segregated in forgotten, ill-supported schools where staff already felt punished enough.  Partly to protect their vulnerable kids, school staff began gaming the numbers, or even outright cheating, to avoid further demoralization.

Eventually, increasing amounts of school time were devoted to test prep, effectively passing the pain on to the kids.  Neither the feds, states, nor researchers recommended becoming test-prep factories.  Sure enough, test-prep barely budged academic performance.  But many schools argued that they had no choice but to focus on the test results because of perceived threats to their jobs.  Then situation made the parents crazy.  All parties blame the tests themselves.  And here we are.

Exactly who is responsible for kids’ learning?

Currently Finland’s high-performing schools are Education’s darling.  Interestingly, their students are tested constantly, but to good effect.  As Anu Partanen writes in a recent Atlantic piece about Finnish schools, “teachers are trained to assess children in classrooms using independent tests they create themselves. All children receive a report card at the end of each semester, but these reports are based on individualized grading by each teacher.”  At the end of high school, students must pass nationally-mandated high-school exit exams, so there is a goal and objective measure of success at the end.

How schools get there is their business.  The Finnish feds regularly check up on schools across the country with tests, but their purpose is to make sure the schools and kids are doing well.  It’s not in their culture to think that the way to produce improvement is to get all nasty at schools that struggle.

On the contrary, the Finns have the attitude that testing provides superb data to help teachers collaborate with one another, along with the child and family, to ensure each kid’s success.  Teachers have enormous responsibility for student learning, but they’re not alone; they lead a circle of adults who own that child’s success, as defined by observation as well as objective data.

By all means test, and even publish the results.

But for the foreseeable future, the feds and states need to rethink their get-tough relationship to school improvement.  NCLB made it painfully clear schools can’t be harangued, threatened or sanctioned into improvement.  (People can’t either.)  So for now, collect data on the kids’ achievement.  And publish it.  But then study it carefully, discuss it, take responsibility for it.  Make sure the data becomes useful wisdom, and quit using test scores as billy clubs.


Julia Steiny is a freelance columnist who also blogs about Restorative Practices and Restorative Justice. After serving on the Providence School Board, she became the Providence Journal’s education columnist for 16 years, and has written for many other outlets. As the founding director of the Youth Restoration Project, she’s been building demonstration projects in Rhode Island since 2008. She analyses data and provides communications consulting on Information Works! and the RIDataHUB, through The Providence Plan. For more detail, see or contact her at or 24 Corliss Street #40022, Providence, RI 02904.

, , , ,

Leave a comment

Will Bureaucratic Evaluation Systems Nourish Great Teaching?

Published by — A rabid commitment to overhauling teacher evaluations is no panacea.  Lessons from NCLB and union involvement are key.

Stop. Take a deep breath.

What do we hope to accomplish with America’s new fanaticism for teacher evaluation?

Other than ridding ourselves of the small proportion of truly wretched teachers – at long, long last! – will these evaluation systems promote excited, avid teaching and learning? Will they rally public support for teachers and their work?

Or will they be yet another exercise in using data to enforce compliance, like No Child Left Behind (NCLB)?

NCLB was a howling success at pushing states and districts to build robust and useful data systems. But then it used the data to focus on failure. It did not disseminate examples of kids and teachers working well and HAPPILY together and, by the way, kicking butt on the numbers – the tests, attendance and graduation rates. It did not promote vibrant, engaged teaching.

Given America’s generally simplistic and punitive mindset, we tend to use numbers as bludgeons. In a moment I’ll suggest how we might make the numbers more useful. But first let’s remember that NCLB’s emphasis on meeting-the-numbers-or-bust led to an epidemic of testing scandals. And its achievement gains were impressively modest considering the money and angst that went into getting them.

Evaluating teachers seems to be school reform’s new silver bullet. The harsh light of punitive “accountability” is turning to individual teachers.

Mind you, teacher evaluations and data systems are both critical to improving education. Teachers need and deserve rich feedback on their work so they can promote and model life-long learning. Data helps them and their colleagues confirm the fruits of their labor and flag points of weakness.

But the most recent effort, the federal Race to the Top grant process, pushed states to create evaluation systems in which student test scores often count for as much as 50 percent of the evaluations. NCLB set the absurd goal of having all kids 100 percent proficient by 2014. Is this teacher-evaluation 50 percent similarly realistic? Won’t these numeric targets for individual teachers just add more heat to the boiling crock-pots that so many frustrated, struggling schools already are?

Furthermore, districts are turning themselves into pretzels trying to apply the state test scores to evaluations of teachers who do not teach tested subjects or grades. “Fair” evaluation systems must apply the rules equally to gym, music, 4th grade and biology teachers. What a lot of work, and for what?

Public relations teams in districts and states are frantically asserting that these new evals will be “formative.” They’ll provoke rich conversations about teaching. They will not be witch hunts.

I’m dubious. The numbers will get in the way.

Because the real problem of public school evaluations is devising a system that will stand up to a court challenge. No matter how incompetent, teachers have every right to insist their union fight for their jobs. If the case goes to court and the district loses, the whole system is shot down. So just to make evaluations minimally viable, they must be based on objective, verifiable, unquestionable data.

So here’s what could happen. Since district administrations are the ones with the data systems, they could generate crisp, easy-to-read data analyses about problem teachers and ask nicely that the unions exercise some quality control.

For example, a GoLocalProv analysis of the Providence School Department data revealed that 37 percent of the teachers were absent 19 days or more. That’s over 10 percent of a 180-day year. Furthermore, 11.5 percent were out twice that much time. Start there.

Certain kinds of bad teachers are easily flagged with readily-available data. Union officials could quietly point out that the data strongly indicate that the teacher’s not really into his job. Maybe there’s a story behind the data. Maybe classroom mold has been making a certain teacher sick a lot. Okay, unions are generally good at helping teachers solve those kinds of problems. But if they find a teacher who persists in abusing sick leave, they could explain that the data make it unwise or unethical to expend union dues protecting him or her.

Think: the American Medical Association and the Bar Association maintain standards for the profession and weed out those who ignore them.

If unions were the ones to show up asking questions about non-controversial data like absenteeism, the ranks of the obviously bad would thin quickly.

The public would applaud. Achievement might rise.

And the stupid reasons for poor teaching would be addressed quickly and discretely. The unions could start shedding their reputation for protecting incompetence.

Union help with removing dead wood would go a long way to defuse the adversarial standoff that bogs down many of our schools. Most importantly, it would free ALL parties for a cooperative conversation about what makes terrific teaching. When the meaning of “quality” is more clearly fleshed out, and adapted to each kind of teacher, THEN we can work on evaluations that will help us be absolutely sure all kids are getting what they need from us.

Without bludgeons.

So much of teachers’ daily frustration concerns non-compliance among the kids. They’re ill-mannered and disengaged. But brow-beating teachers into compliance with numerical goals threatens to make them as surly and uncooperative as those kids. It’s no way to nourish teachers stoking students’ appetite for learning.

Which should be the point of teacher evaluations.

Julia Steiny is a freelance columnist whose work also regularly appears She is the founding director of the Youth Restoration Project, a restorative-practices initiative, currently building a demonstration project in Central Falls, Rhode Island. She consults for schools and government initiatives, including regular work for The Providence Plan for whom she analyzes data. For more detail, or contact her at or c/o GoLocalProv, 44 Weybosset Street, Providence, RI 02903.



, , ,

Leave a comment