Nonprofit, nonpartisan journalism. Supported by readers.

MinnPost's education reporting is made possible by a grant from the Bush Foundation.

Minnesota hopes to avoid Chicago’s pitfalls with teacher evaluations

REUTERS/John Gress
One of the thorniest issues for striking Chicago teachers is the Chicago Public Schools' plan to tie teacher evaluation to students' test scores.

Like a growing number of states enticed by the prospect of federal grants, Illinois and Minnesota both have new laws on the books mandating the creation of teacher evaluation systems that take student performance data into account.

In Chicago, the proposed evaluation system is behind the thorniest of the impasses in the week-old teachers strike. The tests, Chicago Teachers Union leaders insist, will not measure teacher effectiveness yet will cost up to 6,000 teachers their jobs over the next two years.

The angriest among them have accused Chicago Public Schools of wanting to shed the jobs as part of a move to close dozens of schools and funnel the displaced students into non-union charter schools. Not true, CPS leaders counter, noting that for decades virtually all teachers have been rated satisfactory.

Minnesota teachers nervously watching

Minnesota teachers, quite understandably, are watching nervously. By 2014, they will be evaluated annually, with student test results making up a third of their score.

The evaluation systems called for in both places are built around so-called value-added tests, which measure how much a student learned during a given year, and in theory anyhow, by extension can identify the most and least effective teachers.

So does this mean the Chicago strike is a harbinger of what could happen here?

No, education policymakers, district leaders and union officials agree. So far, Minnesota has avoided the potholes that ignited the issue in Chicago.

The biggest difference is the stakes, said Dave Heisted, director of research, evaluation and assessment for Bloomington Public Schools and a member of the state task force assigned to create an evaluation system Minnesota districts can adopt.

Under the system proposed in Chicago, underperformers could be fired after two subpar evaluations. Teachers in areas where standardized tests are not given will be evaluated on building-wide scores. And if union brass are to be believed, no accommodations will be made for influences such as student poverty, special ed status or ability to speak English.

Brenda Cassellius

In Minnesota, the task force has recommended to state Commissioner of Education Brenda Cassellius that the evaluation system under construction be separate from decisions about tenure, compensation and firing.

“There’s no proposal to hire or fire on the basis of value-added tests,” said Heisted. “Our recommendation to the commissioner is to keep to low-stakes use of data.”

Past evaluations inconsistent

In the not too distant past, schools were woefully inconsistent about evaluating teachers. Frequently they were assessed at the start of their careers when their status was probationary but rarely, if ever, after that. Those evaluations that have taken place have too often been a single short “drive-by” observation by the principal, who often knew little about discerning effectiveness.

And the student tests mandated by the federal No Child Left Behind Act only added a layer of paranoia. Schools and districts were judged to be failing if too few students scored as proficient, giving administrators and politicians a cudgel to wave at teachers who didn’t bolster the bottom line.

The advent of the value-added tests, which measure an individual student’s growth over time, promised both to help teachers pinpoint kids’ skills gaps and to make it possible to identify teachers who consistently achieve more — or less — than a year’s progress at a time. Barack Obama and his education secretary, onetime CPS head Arne Duncan, offered cash-strapped states big incentives to start using the data.

The best use of the data, in Heisted’s opinion and that of other local assessment experts, is to identify both student skills gaps and highly effective teachers whose practices should be emulated. Teachers will need time to come to see the data stream as a tool for growth instead of a punitive weapon.

To that end, Minnesota is moving more slowly than Chicago in part to give educators time to see that properly implemented evaluations can be used to help teachers improve their performance, he said.

Illinois lawmakers voted in 2010 to require that student achievement data make up at least 30 percent of teacher evaluations by the 2016-2017 school year and to streamline the process of firing underperformers. When Rahm Emanuel was elected mayor of Chicago in January of 2011, he announced plans to roll out the evaluation in his city this fall, four years early, and to base 40 percent on test scores.

A few months later, Minnesota lawmakers voted to require 35 percent of evaluations, made mandatory for the first time, to be based on student performance. Because the Legislature and Gov. Mark Dayton deadlocked on all but the biggest policy prescriptions, a task force was appointed to work out the details.

In addition to Heisted, the panel is populated by teachers, administrators, union leaders, parents and others with a broad range of views. Minnesota school districts are free to come up with their own evaluation systems provided they fulfill the law, or they can use the system it is slated to recommend in January.

Consistency a key for Minnesota group 

The group is likely to insist that unlike the Chicago model, all evaluators receive training to ensure consistent results and that all teachers, not just underperformers, set goals for professional development and document their progress. Also, it is likely to want mechanisms to assess student growth in non-tested subjects, such as music or phys ed, so that teachers are evaluated on their own work and not their peers’.

Whatever its final recommendation, the Minnesota group’s proposal is likely to be much more nuanced than the one at issue in Chicago. Heisted and the Bush Foundation have been working with the Madison-based Value Added Research Center for several years to find meaningful ways to tie student data to districts and schools, individual classrooms and teachers and even the teacher’s training program.

In the process, assessment experts here have figured out how to adjust the data to account for such variables as student poverty, special education status and English-language ability. So far, they have learned that accumulating information for three years makes it possible to identify outstanding teachers whose practices merit study.

Crucially, in some districts the teachers union has been in on the whole process. This will be the second year where Minneapolis teachers are observed by trained, certified assessors using a series of standards of effective teaching developed with the Minneapolis Federation of Teachers.

The observations were piloted by 1,000 teachers last year, with the MFT soliciting feedback that was used to tweak the process going forward. This year, all teachers will be observed and given their own students’ test scores to study in private [PDF]. Administrators will only have access to school-level data.

Students will be surveyed, and their opinions shared with teachers only. If things go the way MPS leaders hope, teachers will decide to compare notes and share useful strategies, said Maggie Sullivan, administrator of strategic planning.

“We really want to make sure this is useful to teachers,” said Sullivan. “The MFT has been a great source of feedback for us because they hear from a lot of teachers.”

Trust issue important

Lynn Nordgren

And a lot of what they hear, according to MFT President Lynn Nordgren, is fear of a “gotcha model” and hunger for meaningful feedback. “It’s partly an issue of trust,” she said.

“We’re interested as teachers in having this flattened out so it’s not top down,” said Nordgren.  “We know that if we have a model that’s not based on fear, people will be more accepting. In order to get there, we need multiple means to get that data.”

So far, principals have had more struggles than teachers with the evaluation model under construction in Minneapolis, she added. The sheer number of hours required to conduct good evaluations is one issue.

Nordgren is concerned that MPS’ and Minnesota’s collaborative works-in-progress could still become gotcha models. “It’s not a done deal,” she said.

Helen Fisk is head of Global Academy, a high-performing Columbia Heights charter school where she can hire and fire at will. “How our students do on standardized tests is, of course, very important to us,” she said. “But I actually haven’t ever had to have a conversation with a teacher about test scores.”

Her teachers meet in teams and talk about the scores as part of an ongoing conversation about student performance. But low scores are never the first sign that there’s a problem with an instructor.

“If a teacher has trouble getting kids to perform well on standardized tests, there are other issues,” she said. “They can’t manage their classrooms, they have discipline problems. It’s kind of a package deal.”

Comments (9)

  1. Anonymous Submitted by Anonymous on 09/14/2012 - 09:48 am.

    This is all complete baloney. The student tests were not designed to rate teachers – see the research. The test makers should be up in arms about this, but they’re not, because it means big bucks to them. Second, the so called Value-Added models are completely useless. They have stated margins of error of 35 points, so a teacher could have a 15 rating one year and 85 the next.

    These ratings of teachers based on student test scores are even more useless without multiple years of scores for each teacher. Given our new “alternative licensure” measures, i.e. the Teach For America Enabling Act, these measures will never mean anything to the Teach for Awhile teachers from TFA, because upwards of 80 percent of them are gone within three years! And I haven’t even gotten into both the outside factors that comprise the educational experience, i.e. poverty, and the disservice that standardized testing has on narrowing curriculum, harming critical thinking abilities, and rampant teaching to the test.

    Don’t believe for a minute that just because advocates say these testing laws won’t be egregiously misused in the onset of their use they won’t be later.

  2. Submitted by Jim Bartholomew on 09/14/2012 - 10:30 am.

    What do we expect?

    As part of the 35% of teacher evaluations, are we really willing to settle for expecting all students to make one year’s worth of “normative” growth?

    Sure, one could argue this may be fine for students who’re already meeting or exceeding grade level standards/expectations. However, for kids who are performing below expectations, we either need to set higher expectations, or be up-front and say eliminating achievement gaps isn’t an urgent priority.

    Put another way, if the expected amount of yearly growth for low-income students (or students of color) allows them to continue performing below grade level – with no set time frame for reaching grade level – what’s the message?!

    We can and should recognize teachers who’re successful in helping all students meet expectations. We should also look at how to provide incentives for these teachers to work with “struggling” students.

    • Anonymous Submitted by Anonymous on 09/14/2012 - 10:55 am.

      The “incentives” go just the opposite way – why should teachers go into high-poverty schools when it is highly likely that no matter what they do their students won’t pass the tests? Why not move to middle or upper class schools where the students will do well? And ask yourself this: Just WHERE did the 35% number come from? Certainly not from educational research. It is a politically derived number pulled from education deformers’ rear ends.

      • Submitted by Thomas Swift on 09/14/2012 - 11:30 am.

        A better question…

        “why should teachers go into high-poverty schools when it is highly likely that no matter what they do their students won’t pass the tests?”

        How about why should the public pour billions of dollars into a failed government system when it is highly likely that no matter what they do their students won’t pass the tests?

        The only people prospering in the status quo are carpet bagging district superintendents and the NEA. Thankfully, we are rapidly nearing the end of the line for the gravy train express.

        • Submitted by Ross Reishus on 09/24/2012 - 06:56 pm.

          the answer is not so simple

          “why should the public pour billions of dollars into a failed government system when it is highly likely that no matter what they do their students won’t pass the tests?” (BTW I don’t hear conservatives screaming about our military’s lack of “effectiveness” in Afganistan, yet we spend 10 times the $ there. Another topic for another day.)

          No test of any kind will properly measure the work of a teacher who has students who’s parents are a. beating each other up and possibly the child, or b. abusing drugs/alcohol, or c. have lost their home due to greedy gotcha-motive banks, or d. mentally ill, or e. have no jobs, or f. have to work 4 of them to pay bills, OR a. and b., b and c., c. and d., or d. and e. or any other combination of a,b,c,d,e,f, or even all of the above……

          When any of these circumstances exist…
          1. Its difficult for the child to take in what’s going on in school.
          2. Its not the child’s fault.
          3. Its also not the teacher’s fault.
          4. Now ad some language barriers, and all the lack of funding that goes with that.

          Who to blame? On the government side….In addition to continued reductions in education funding, social serivices and community outreach group funding has been repeatedly slashed both at the state and federal levels since the 90’s. In other words, several of our social safety nets are either missing, or have gaping holes in them. Let’s elect some people who will repair the damage. On the private sector side there is much to blame. Who sent all the jobs overseas? Greedy CEO’s and their boards mostly. They make those decisions, and their boards pay them embarrassingly absurd salaries to do it. This makes it hard on many middle class and struggling class families. Who keeps pushing for further deregulation and fewer consumer projection laws? Usually the same people who are pushing for education cuts. An informed consumer is not a very desirable demographic for a lot of businesses. All of these things are factors, believe it or not.

          And yes, if this absurd test scores = teacher effectiveness idea makes it impossible to keep your job by working in an economically challenged urban school system then no one will work there and now you have even bigger problems.

          To summarize, its not a simple problem with a simple solution, in spite of what a.m. talk radio would like you to think.

  3. Submitted by Beth Hawkins on 09/14/2012 - 11:59 am.

    Jim Bartholomew

    Jim, I’m going to point out for readers who may not know that you are the education policy director for the Minnesota Business Partnership, a member of the state Teacher Evaluation Working Group referred to in this story and a past member of the state Board of Teaching.

    And I wonder whether you have been following our series on school culture? I’ve been spending time in schools that are expecting much more than one year’s normative growth–some have kids that are entering high school five-plus years behind. And there are teachers who are intentionally bidding into them, in the case of Minneapolis’ North, or who are TFA alums who want to continue teaching because they want to be in those settings. Two of those schools rely on frequent formative assessments to “scaffold” students up much more rapidly than a year at a time. Those teachers expect to be assessed, formally and informally, much more often. They also have very supportive school structures, work in professional teams and do not struggle with class sizes anywhere near those of their mainline urban counterparts.

    • Submitted by Jim Bartholomew on 09/14/2012 - 05:55 pm.

      It begins with expectations

      Hi Beth: Thank you, I have followed your series, and in many ways the discussion of how much academic growth we expect for students is the fundamental question for education in Minnesota.

      Since 1985, I’ve been involved with education policy in Minnesota (and Wisconsin), I serve on the Board of Directors of one of the two other schools you highlighted and the MN Business Partnership has been recognizing high-performing, high-poverty public schools since 2006.

      A common theme among successful schools is they firmly believe each student can and will meet/exceed grade level expectations. When students who’re performing below grade level come to these schools, the schools don’t expect them to make just one year’s worth of academic growth, they expect more.

      As you point out, Minnesota has schools and teachers that are closing achievement gaps and helping students meet/exceed expectations. We should be sharing the practices and expectations of these teachers and schools, not adopting measures in teacher evaluations that are based on students making a year’s worth of growth when they need more.

      One of the five education gaps we need to close is the Belief gap – and this begins by believing students, regardless of race or income, can meet and exceed grade level standards.

      We have a tremendous opportunity to help students and teachers through the development of an effective evaluation system that aligns our education system with the goal of all students graduating prepared for a post-secondary education or entry into a career.

  4. Submitted by Ray Schoch on 09/14/2012 - 01:29 pm.

    Mr. Swift is amusing


    His tone suggests that public schools are somehow the product of some foreign entity he labels “government,” when public schools are, in every single case, the product of – the hopes and dreams of – the local community. There simply are no widespread entities in this society that enjoy more local control and that get more local input, both positive and negative, than public schools.

    Readers should note, as well, that Mr. Swift provides no credible alternative to the system he labels as a failure, though he either never experienced it himself, or else owes the foundation of whatever success in life he’s enjoyed to the same system he now condemns. It’s an interesting intellectual posture… Meanwhile, one would think that someone with Mr. Swift’s highly developed entrepreneurial skills would have gleefully climbed aboard any “gravy train” as lucrative as he implies public education to be.

    Rating teachers on student performance on a standardized test doesn’t make a whole lot of sense to me in the first place. When it’s a test in which those students have zero personal investment it amounts to lunatic public policy, for all the reasons Rob Levine has provided, plus a few more. The only way I would support standardized testing as a measure of anything of significance would be if the students themselves had some investment in the outcome. As it stands, a kid who aces the test gets no meaningful reward, and the otherwise-capable kid who uses his answer sheet as a pincushion and scores in the 20th percentile as a result suffers no negative consequence for doing so.

    “Struggling” students come from poverty, or otherwise terrible home situations, over which teachers have no control. Or they have serious physiological or emotional handicaps over which, once again, the teacher has no control. Sometimes, they come from both sets of circumstances. Yes, there are teachers who sometimes can work miracles with those kinds of kids – I worked a few of those miracles myself over my years in the classroom – but miracle-working on a regular and frequent basis is not a realistic expectation for a human being, else we’d all be walking on water and using “Saint” as a prefix for our names.

    Some teachers are better than others, it’s true. Some connect better to their students, at whatever age/grade level. You want to pay them more for that ability to engage children in learning? Fine. Just don’t base that pay scale on test scores that mean nothing, or on student “growth.” There’s no way to assess with any accuracy how much of said student’s “growth” is attributable to her experience in my classroom, compared to her experience outside of school, or even to her experience while engaged in her own, private, search for wisdom on her own terms. If you’re going to penalize me for the kid whose standardized test score is abysmal, what sort of bonus will I get for teaching that Rhodes Scholar, or the kid who gets a full-ride academic scholarship to Stanford or Harvard?

    Another point – this doesn’t seem to appear in the comment threads very often – is this: If we’re going to turn evaluation into the sort of “competitive” environment that Mr. Swift apparently relishes, why would a teacher who’s “successful” in those terms, whose students regularly garner top-notch test scores and show above-average “growth” during their year with her, want to share her techniques and tips with her fellow teachers when doing so might enable her colleagues to end up being paid more than she’s being paid because their students manage to get higher test scores than her students? Once you “incentivize” the process by basing evaluations on student scores, not many people are going to be willing to share what works for them if they know that the result might be that someone else gets the “performance bonus.”

    There are very good reasons why Lynn Nordgren and her MFT colleagues are suspicious of a “gotcha” model. Every teacher wants meaningful feedback, and the people I worked with for 30 years were always interested in a more effective way to present a concept or piece of narrative, but as she’s quoted as saying, “It’s…an issue of trust.”

  5. Submitted by Lynnell Mickelsen on 09/14/2012 - 02:00 pm.

    Great reporting, Beth

    It’s a complicated topic. Over the last couple of years, I have learned to be more patient when it comes to developing teacher evaluations, precisely for all the reasons in your story.

    I’ve gone from wanting to start something as in right now….to seeing the value in taking time to get the evaluations right. I agree with the MFT’s Lynn Nordgren that teachers will be more accepting of a model that’s not based on fear. That’s why the St. Francis model is so appealing.

    And of course, I support using multiple measures—i.e. not just test scores. Frankly I don’t know any experts who want to evaluate teachers based only on standardized test scores, so it’s a straw-man argument.

    One reason teachers are so down on test score data being used is that they spent years watching data being badly—even malevolently—used under No Child Left Behind to the point where a lot of them have a PSTD reaction to it.

    I find that when value-added data is explained, a lot of teachers are more open to it. I also agree that value-added data should be demographically-weighted because I don’t want to repeat the NCLB pattern of punishing teachers who chooses to work with low-income students of color. At the same time, I understand the legitimate concerns of Jim Bartholomew and others about lowering the performance bar for poor students.

    I still don’t have an answer about the best way to evaluate art, music and gym teachers. Jimmy Barnhill of the MFT has also schooled me on how especially difficult it will be to evaluate special ed students with data—I’ve learned a lot from him.

    It will be interesting to follow what happens in Indiana, where they are launching statewide teacher evaluation this year and I believe they plan to use the results in hiring decisions starting the next year.

Leave a Reply