examining gaps

At-risk students in some big Colorado districts have a better chance of having an effective teacher than others

PHOTO: Denver Post file

While at-risk students statewide were less likely to have teachers rated effective or higher in the 2014-15 school year, the gaps between the percentage of effective teachers in high-poverty and low-poverty and high-minority and low-minority schools varied greatly by district.

In Denver Public Schools, the state’s largest school district, the gaps were even wider than state averages, according to data released this week.

Students in the second-largest district, Jefferson County, experienced the opposite: Kids in high-poverty and high-minority schools were actually more likely to have an effective teacher.

And in Douglas County, large gaps that showed poor students at a significant disadvantage in terms of teacher effectiveness were attributable in part to the fact that the wealthy district has so few high-poverty schools and so many low-poverty schools.

The data was released nearly seven years after state lawmakers adopted a landmark teacher and principal evaluation system. The 2014-15 school year was the first year it was fully in effect. Under the law, the state must look at the number of effective educators in schools that serve varying levels of low-income students, students of color and English language learners.

To calculate those gaps, the state education department ranked every school in Colorado from highest to lowest by the percentage of students in each of those three groups.

The state then broke each of those lists into four quartiles and compared the percentages of effective teachers in schools in the highest and lowest quartiles for each district.

The gaps in Denver were bigger than statewide averages.

For example, in DPS schools with high proportions of English language learners, 62 percent of teachers were rated effective or higher, the data show. In schools with low proportions of English language learners, 89 percent of teachers were effective or better.

That’s a 27 percentage-point gap. Statewide, the gap was 8 percentage points.

The numbers were almost exactly the same for DPS schools with high and low proportions of students of color: 63 percent versus 89 percent, respectively.

And in DPS schools with high proportions of students living in poverty, 64 percent of teachers were rated effective or higher. In low-poverty schools, 84 percent were effective or better.

DPS uses its own teacher evaluation system, which meets Colorado requirements but is different than the state-developed system most school districts use. Denver also had a high percentage of teachers show up in state data as “not rated” in 2014-15 for a variety of reasons related to attrition, new hires and the large number of charter schools in DPS.

But while DPS officials said they’re still digging into whether the state’s gap analysis lines up with the district’s own number-crunching, they acknowledged that DPS is “not satisfied where we are” when it comes to teacher effectiveness gaps.

“One of the things we have been focused on is getting our strongest teachers to come to and stay in some of our highest needs schools,” said Sarah Almy, the district’s executive director of talent management. “…One of the challenges — and one of factors in that gap — is that we, as many districts do, struggle to retain teachers in our highest needs schools and consequently wind up with a greater proportion of new teachers in those schools.”

New teachers are more likely to be rated “partially effective” — as opposed to “effective” or “highly effective” — than veteran teachers, Almy said. DPS has been trying to attract more effective teachers to high-needs schools by offering them financial incentives, she said. This year, the district is also focusing on increasing teacher retention in those schools.

In neighboring Jefferson County, state data show 90 percent of teachers in schools that serve the county’s poorest students were rated effective or higher. At the same time, 82 percent of teachers in the county’s wealthiest schools earned one of the top two ratings.

Similar inverse gaps existed in schools that serve high and low proportions of English language learners and students of color.

Todd Engels, Jeffco’s executive director of educator effectiveness, said the district is studying the data but noted it could be difficult to draw any conclusions given how old it is.

“We’re thankful that we have some great teachers in those high-needs schools,” he said.

One possible reason for the reverse gap, Engels said, is that a dozen of the district’s highest poverty schools have been involved in a national pay-for-performance study known as Strat Comp to test new ways to pay teachers and identify what helps them become better instructors.

Teachers in the study were evaluated by both school administrators and trained peer evaluators that worked across multiple schools. Some teachers received bonuses up to $15,000 tied to their evaluations, while others received stipends. All teachers received a higher starting salary.

The research was funded by a $32.8 million five-year grant from the federal government. Engels said the district has not yet received the final report on the impact of the investment, but participating schools have been flushed with additional resources and training for teachers and principals.

One of the largest gaps based on poverty in the state was in wealthy, suburban Douglas County. In high-poverty Dougco schools, 42 percent of teachers were rated effective or higher. In low-poverty Dougco schools, 79 percent were effective or higher.

But according to the way the state calculated the gaps, the district had 79 schools in the wealthiest quartile and only three in the poorest quartile. All three are run by the HOPE Online charter organization, a multi-district online school with learning centers along the Front Range.

Britt Wilkenfeld, director of research for educator talent at the Colorado Department of Education, said that because of the exceedingly low number of Dougco schools in the high-poverty quartile, the gap analysis there “might not be as meaningful” as in other districts.

“You’re really just looking at the gap between that school and the rest of the district,” she said.

Douglas County School District officials did not provide responses to Chalkbeat questions by the end of business Tuesday.

Incentives

Westminster district will give bonuses if state ratings rise, teachers wonder whether performance pay system is coming

PHOTO: Nicholas Garcia
Students work on an English assignment at M. Scott Carpenter Middle School in Westminster.

Teachers and employees in Westminster Public Schools will be able to earn a bonus if they help the struggling district improve its state ratings next year.

The district’s school board on Tuesday unanimously approved the $1.7 million plan for the one-year performance stipends, the district’s latest attempt to lift the quality of its schools.

School employees can earn $1,000 if their school meets a district-set score, or up to $2,000 if they reach a more ambitious goal the school sets. District employees, including the superintendent, can earn $1,000 if the district as a whole jumps up a rating next year.

“We recognize that everyone plays a critical role in increasing student achievement and we decided that if a particular school or the district as a whole can reach that next academic accreditation level, the employees directly responsible should be rewarded,” board president Dino Valente said in a statement.

The district is one of five that was flagged by the state for chronic low performance and was put on a state-ordered improvement plan this spring.

District officials have disputed state ratings, claiming the state’s system is not fairly assessing the performance of Westminster schools. Middle school teacher Melissa Duran, who also used to be president of the teacher’s union, drew a connection between that stance and the new stipends, saying any extra pay she gets would be based on one score.

“The district has gone to the state saying, ‘Why are you rating us on these tests, look at all the other things we’re doing’” Duran said. “Well, it’s the same thing for teachers. They’re still basing our effectiveness on a test score.”

Teachers interviewed Thursday said their first thoughts upon learning of the plan was that it sounded like the beginnings of performance pay.

“I already get the point that we are in need of having our test scores come up,” said math teacher Andy Hartman, who is also head of negotiations for the teacher’s union. “Putting this little carrot out there isn’t going to change anything. I personally do not like performance pay. It’s a very slippery slope.”

District leaders say they talked to all district principals after the announcement Wednesday, and heard positive feedback.

“A lot of the teachers think this is a good thing,” said Steve Saunders, the district’s spokesman.

National studies on the effectiveness of performance pay stipends and merit pay have shown mixed results. One recent study from Vanderbilt University concluded that they can be effective, but that the design of the systems makes a difference.

In Denver Public Schools, the district has a performance-pay system to give raises and bonuses to teachers in various situations. Studies of that model have found that some teachers don’t completely understand the system and that it’s not always tied to better student outcomes.

Westminster officials said they have never formally discussed performance pay, and said that these stipends are being funded for one year with an unanticipated IRS refund.

Westminster teachers said they have ideas for other strategies that could make a quick impact, such as higher pay for substitutes so teachers aren’t losing their planning periods filling in for each other when subs are difficult to find.

Waiting on a bonus that might come next year is not providing any new motivation, teachers said.

“It’s a slap in the face,” Duran said. “It’s not like we are not already working hard enough. Personally, I already give 110 percent. I’ve always given 110 percent.”

Last month, the school board also approved a new contract for teachers and staff. Under the new agreement, teachers and staff got a raise of at least 1 percent. They received a similar raise last year.

teacher diversity

Efforts to ‘raise the bar’ for becoming a teacher are running headlong into efforts to diversify the profession. Now what?

PHOTO: Meghan Mangrum

Education advocates and policymakers want to have it both ways: they want more teachers of color and to “raise the bar” for the profession with measures that disproportionately screen out certain groups.

The two aims, both widely popular in the education policy circles, aren’t just on a collision course. They’ve already collided. In Baltimore, for instance, a highly-rated black teacher may lose her job because of a licensing exam.

This the third story in a three-part series on the relationship between certification and teacher diversity in America. Read part one and part two.

But there has been only limited discussion of the fact that these two objectives — diversifying the profession and making it harder to enter — are often at cross purposes, although certification rules are hardly the only reason for limited diversity among teachers.

“You need to think pretty comprehensively if you’re going to accomplish both those goals,” said Dan Goldhaber, a professor at the University of Washington and a leading researcher on teacher certification. “Some of the things you do to accomplish one goal work against the other.”

A report released Thursday by the Center for American Progress, a progressive think tank, acknowledges the tension but argues that it’s possible for schools to have it all.

It suggests states confront the issue on multiple levels, investing more in recruiting potential teachers of color while also making the profession more appealing to everyone by raising salaries and subsidizing training. All of that should coincide with the development of better tools for judging whether a prospective teacher will be effective, which could replace current requirements like GPA cut-offs, CAP argues.

“Rigorous recruitment and thoughtful selection processes can achieve increased diversity and selectivity simultaneously,” the report says.

The paper offers one blueprint for policymakers. Others say states ought to take the opposite approach and actually lower the bar for entry into the profession, then carefully measure teachers’ performance once they’re in the classroom.

But right now, there isn’t much obvious political will to implement any of it, or clear research on the best approach.

Some states really have raised the bar

Between 2011 and 2015, nearly half of all states have ratcheted up testing or GPA requirements for entering teacher training programs. The raise-the-bar message has become policy, and teachers of color are the most affected.

At the same time, the push to increase the diversity of the overwhelmingly white teaching force has grown more urgent in the wake of recent studies.

“There’s clear qualitative and quantitative research that points to the added value for students of color when taught by a teacher of color,” said Travis Bristol, a professor at Boston University.

States have created an array of task forces to figure out how to recruit more teachers of color; the federal government has repeatedly made the case for doing so. Think tanks and policy groups have issued reports and held panel discussions. Many of those same groups have also called for raising the bar to enter the profession.

Meanwhile, states continue to grapple with how many hoops teachers should have to jump through before reaching the classroom. New York recently eliminated one exam required to become a teacher largely because of concerns about its impact on diversity, and is also considering dramatically reducing requirements for teachers at certain charter schools.

CAP offers a variety of solutions

The CAP paper points to a handful of states and teacher training programs that have prioritized both teacher diversity and high standards, like the Boston Teacher Residency.

It concludes by recommending improving recruitment for teachers of color, increasing teacher pay, using multiple measures for evaluating prospective teachers, and researching better metrics for predicting teacher effectiveness.

Goldhaber says a comprehensive approach is necessary.

“If you only [raise the bar], that’s going to be probably harmful to workforce diversity,” he said. “If you do that in connection with some other things, like raising salaries and reaching out to different kinds of people — those initiatives in conjunction with one another could work to both increase the diversity of the workforce and raise the bar.”

Of course, many of these ideas, like paying teachers more, cost money. As of 2014, many states were spending less on education than they were prior to the Great Recession.

Finding better measures for predicting who will be successful in the classroom may also be challenging, as existing metrics have proven limited and there isn’t a consistent definition of quality teaching.

Some research suggests that using a combination of metrics may be a useful tactic. Catherine Brown of CAP says teachers should be judged on a greater variety of skills without as much weight put on one test.

“One single test is not enough to tell you who will be a good teacher,” she said. “You need to look at the entirety of the candidate and that includes their interpersonal skills, and their ability to control a classroom, and their cultural competency with students.”

PHOTO: Shaina Cavazos

New tests and new kinds of training

Some are optimistic about a new breed of exam, including the edTPA, for directly measuring teachers’ effectiveness and reducing racial disparities in performance. The edTPA aims to examine a teacher’s practice, and it shows smaller racial gaps than paper-and-pencil certification tests — though substantial differences still exist between black and white candidates.

“I think that performance-based assessments [like edTPA] really give us an insight into how someone is able to perform on the job,” said Bristol.

Research on whether edTPA predicts effectiveness is only in its infancy. One analysis in North Carolina had promising results: a score on the performance assessment was a moderately strong predictor of teachers’ ability to raise student achievement, as well as how they were rated by their principals. Another study in Washington state was more measured — the ability to pass the edTPA was modestly related to effectiveness for English, but not math, teachers.

Another potential solution, favored by CAP and Bristol, are teacher residency programs, which offer intensive, practical training and the cost is usually heavily subsidized.

Indeed, research has found that teacher residency programs do attract and retain more teachers of color. But one aspect of the model that makes it especially attractive — generous funding to ensure candidates don’t have to go into debt — also makes it difficult to sustain and scale.

A different approach: lower the bar for entry, raise it for staying

To some critics, a better solution is to simply stop trying to divine which teachers will be effective before they reach the classroom — eliminating the possibility of incorrectly screening out qualified people, including prospective teachers of color.

“The current rules end up preventing a significant number and a disproportionate number of candidates of color from being able to teach kids, in spite of their demonstrated ability to do so in the classroom,” said Dan Weisberg, who argues that existing rules are poor proxies for performance.

Weisberg, whose group TNTP runs fast-track training programs for teachers, argues for observing teachers in action and looking at their impact on student test scores instead.

Whether teacher certification status and the associated tests actually predict teacher effectiveness is the subject of a lot of debate. Some educational economists have floated the idea of essentially eliminating certification requirements and focusing on initial performance instead.

Some studies find that certified teachers, and those who score higher on licensing exams, perform somewhat better in the classroom, though the differences are often quite small. (These studies measure teacher performance by their impact on student test scores.)

One advantage to traditional certification: teachers who go through that pathway are much more likely to stay in the classroom. If a different system increased teacher turnover, that could bring its own negative effects.

Goldhaber is not sure it makes sense to simply eliminate certification rules.

“I think you’d likely get more uneven outcomes,” he said. “There are places where that might benefit school systems … but then there are also places where they would not and might select people who are quite ineffective.”