2017 in review

What we’ve learned: 5 lessons from education research to take into 2018

PHOTO: Shaina Cavazos

Education research comes out faster than most of us can keep up with — and staying up to date gets even harder when advocates on every side claim that the newest study supports their views.

We’re here to help. Here are some of the most important lessons we’re taking away from 2017, thanks to the researchers who do their best to separate fact from fiction. (The typical caveats apply: these are all subject to change based on new evidence, and each study has limitations.)

1. Teacher certification rules can have negative side effects.

There are two big ways that rules about who can and can’t teach cause problems. First, they disproportionately exclude teachers of color, who a bevy of recent studies have shown benefit students of color. High-stakes exams, GPA cutoffs, and traditional training requirements all hit would-be teachers of color the hardest, and there’s no clear solution.

Another downside of existing rules: they can make it hard for teachers to move to a new state. A recent study finds that although teachers are less likely to move between states than many other professionals, perhaps because of challenges in gaining a new license.

This can hurt students, particularly if effective teachers leave education as a result. And it may explain another new finding: that schools near state borders — and thus most affected by teachers unable to move between states — have lower student achievement.

2. Union protections may benefit students.

Teachers unions have long argued that by protecting teachers and bargaining for better pay, they ultimately help students. Research bolstered their case this year.

Most prominently, an analysis found that Wisconsin Gov. Scott Walker’s successful effort to dramatically scale back union power hurt student test scores. Another study, this one from California, showed that when charter schools unionize, students saw larger test score gains. That study wasn’t able to pinpoint why.

Two other studies — one from Louisiana, the other from Michigan — showed that removing tenure protections increased teacher turnover, at least in some schools. Past research has found that turnover usually harms students.

3. Students who stay in voucher programs longer do better.

As U.S. Education Secretary Betsy DeVos pushed private school vouchers into the limelight, critics have seized on recent studies showing that using public money to attend a private school can hurt student learning.

But analyses out of Indiana and Louisiana suggest that students who stick around in private school for three to four years see their scores bounce back after an initial drop. By year four, students in Indiana even made some gains in English. Still, many other students saw test scores drop because of the program, including those who left early and students in younger grades in Louisiana.

The disappointing test score results pushed voucher proponents to focus on their impact on other metrics, like high school graduation or college attendance. One study this year found that Florida students who used a tax-credit voucher were more likely to enroll in — though not necessarily complete — two-year college than similar students who attended public school.

4. State tests provide useful information about how schools affect students. Testing can also have unintended consequences.

One study focusing on charter high schools in Chicago showed that not only did those schools raise test scores substantially, they also helped send more kids to college and to stay there. That was also true of Chicago’s Noble charters, a high-profile network. Another piece of research from this year came to a similar conclusion: students who attended high schools in Michigan that raised students’ test scores also earned higher GPAs in college. At least in these contexts, tests were a meaningful gauge of school quality.

However, we also looked at a study showing that students were (slightly) less happy in the classrooms of teachers who were effective at raising test scores. This suggests that there are multiple dimensions to good teaching — and being good at one aspect doesn’t mean you’re good at others.

Finally, another study highlights the challenges of using tests to hold schools accountable: by focusing on test results starting only in third grade — the first year with federally mandated exams — schools are encouraged to place their weaker teachers in earliest grades. And many schools, at least in Miami, Florida, did just that.

5. We still don’t know much about how to turn around a struggling school.

This lesson may be the least surprising to policymakers. But as states try to help low-performing schools under the new federal education law, ESSA, they have a thin research base to draw from.

The highest-profile study on the topic came at the beginning of the year: a federal analysis of the Obama-era turnaround plan known as School Improvement Grants. It did not have any clear benefits — a finding DeVos has since touted to promote her own favored strategies.

But other studies from this year suggest that the effects of the federal improvement grants varied by place: They appear to have had a big impact in both Ohio and San Francisco, but not in Rhode Island.

New York City has also wrestled with this challenge. Early research has found that the city’s high-profile and expensive effort to help schools by offering wraparound services and other help had produced only mixed results.

Cap and gown

Graduation rates in Michigan – and Detroit’s main district — are up, but are most students ready for college?

The state superintendent had some good news to share Wednesday about last year’s four-year graduation rates: They are at their highest level in years.

What’s not clear is whether new graduates are being adequately prepared for college.

Slightly more than 80 percent of the state’s high school students graduated last year, an increase of about half a percentage point from the previous year. It was news state education leaders cheered.

“An 80 percent statewide graduation rate is a new watermark for our schools. They’ve worked hard to steadily improve,” state Superintendent Brian Whiston said in a statement.

“This is another important step in helping Michigan become a Top 10 education state in 10 years. We aren’t there yet, so we need to keep working and moving forward,” he said.

But statewide, the number of students ready for college based on their scores on the Scholastic Aptitude Test was about 35 percent, underscoring the fact that graduation rate is not necessarily a great measure of school success. Schools looking to raise graduation rates can find ways to make it easier for students to earn credits toward graduation and, unlike some states, Michigan does not require students to pass graduation exams.

The result is that more students are graduating from high school — but might not be ready to do college work.

In Detroit, graduation rates in the city’s main district remained largely steady, with a little more than three-quarters of its students graduating after four years. But the number of students who were ready for college dropped almost a point to 12.3 percent last year. While most students take the SAT in 11th grade as part of the state’s school testing program, that’s an indication students graduating from high school may not have been adequately prepared for college.

The state dropout rate remained largely unchanged at almost nine percent.

Detroit’s main district had the highest four-year graduation rates compared to other large districts, but more district students dropped out of school than in the previous year. More than 10 percent of Detroit students dropped out of high school in the 2016-17 school year, a slight increase from last year, according to state data.

But in spite of steady dropout rates and relatively low college readiness numbers, state officials were upbeat about the graduation results.

“This is the first time the statewide four-year graduation rate has surpassed 80 percent since we started calculating rates by cohorts eleven years ago,” said Tom Howell, director of the Michigan Center for Educational Performance and Information, which tracks school data. “This increase is in line with how the statewide graduation rate has been trending gradually upward.”

Search below to see the four-year graduation rates and college readiness rates for all Michigan high schools.

testing testing

New York has debated ‘innovative’ tests. But what does that actually mean?

In New York, the annual state tests still mean the usual multiple-choice questions and short writing prompts — and that’s not likely to change soon.

State officials recently chose not to apply to join a federal “innovative testing” program, which would have triggered an overhaul of the math and English tests that students in grades 3 to 8 take each year. (They cited the cost and difficulty of rolling out the new tests on the tight timeline required by the program.)

Yet in a state where nearly one in five families choose to boycott the exams — which many say do a poor job measuring students’ learning — pressure remains on policymakers to come up with new and better tests. They appear to have some interest in moving in that direction: The state’s education policy-making body, the Board of Regents, has established a workgroup focused partly on testing, and the state education commissioner has expressed interest in alternative graduation exams and new types of science and social studies tests.

If New York does pursue “innovative” assessments, what might they look like?

To answer that, we found four real-world examples of alternative assessments — two used in New York City, and two from other states. While each has its drawbacks, they show that “testing” doesn’t have to mean shading in bubbles with a No. 2 pencil.

Example 1: Tasks used to evaluate teacher effectiveness (New York City)

Who takes them? Students in all grades, in subjects including math, English and the arts

What are they? Essays and short tasks

As part of New York City’s teacher evaluation system, schools can choose from a menu of assessments meant to measure how much teachers have helped their students learn. Among those “Measures of Student Learning” are tasks that require students to make an argument in a written essay by citing examples from texts they are provided.

Other kinds of “MOSL’s,” as they’re often called, go beyond essays. Teachers can administer “running records,” where they assess individual students’ English skills as they read a series of texts out loud. A visual-arts assessment asks students to draw a still-life picture.

What do they look like? A 12th-grade English test from several years ago asks students to answer the question, “Should individuals enlist in the military and fight for their country?” In order to make their case, students must read and cite a poem and a portion of President Woodrow Wilson’s address to Congress.

What are the drawbacks? They are similar to standardized tests

The city teachers union negotiated with the city to include these tests in teacher evaluations, yet many teachers and school leaders have complained that they take up too much class time. Others question whether the assessments fairly measure students’ ability in subjects like art.

“Art is subjective,” said Jake Jacobs, a middle-school art teacher at Bronx Park Middle School. “If somebody is drawing something, who’s to say whether that drawing is good or bad?”  

Example 2: Projects required for graduation (New York City)

Who takes them? Students at 38 New York City high schools in the “Performance Standards Consortium”

What are they? Months-long projects

The consortium was established in the 1990s to bring together New York City educators seeking an alternative to traditional standardized tests. Today, schools in that group have state permission to substitute long-term projects for several of the Regents exams that students must pass in order to graduate. At those schools, each student must complete a literary essay, solve a complex math problem, design a science experiment, and complete a research paper in order to earn a diploma.

In their junior or senior years, they present their projects to a panel of judges, who evaluate whether the work meets the consortium’s requirements.

Sample questions: According to consortium materials, one student wrote a paper titled, “What Role Do Black Characters Play in Faulkner’s The Sound and the Fury and Flannery O’Connor’s Short Stories?” Another produced a science experiment called, “How Does a Garter Snake Detect Its Prey?” Another wrote a research paper titled, “Who Or What Is Responsible for the End of Slavery in the United States?”

What are the drawbacks? Heavy workload for educators and students

Schools in the consortium must spend a lot of time training teachers to oversee students’ projects and ensuring that the work is held to the same standards across the consortium. For that reason, even Ann Cook, executive director and co-founder of the New York Performance Standards Consortium, says it would be difficulty to spread this model across the entire state.

“Could every kid in the state be doing this? I’d probably say no,” Cook said. “And the reason is that you have to have a faculty that is interested and wants to do it. They should want to opt into this because it takes a lot of work in the school.”

Example 3: Real-world problems (New Hampshire)

Who takes them? Students in grades K-12 in math, English, and science

What are they? Tasks based on real-world scenarios

New Hampshire students take traditional standardized exams once in elementary, middle, and high school. In the other years, they take alternative assessments after they finish studying units tied to the state standards. Called “Performance Assessment of Competency Education,” or PACE tests, they challenge students to apply skills they learn in class to real-world problems.

Sample question: After learning how to calculate volume, high school geometry students are given a task where they assume the role of a town planner. Their job is to design a water tower that can hold enough water to support the town’s growing population, but which requires a limited amount of material to build. The project, which students typically finish in a couple hours, must include a cover page, models or scale drawings, the calculations students made, and a written analysis of their design.

What are the drawbacks? Costly and difficult to scale

Though New Hampshire started experimenting with the “PACE” tests more than three years ago, they still have only been rolled out in a fraction of the state’s school districts. The state also had to rely partly on philanthropic funds to develop the expensive assessments, according to Scott Marion, executive director of the Center for Assessment, who helped New Hampshire with its testing experiment.

Because of their high cost and difficulty to roll out statewide, Marion advised New York against adopting similar tests.

Example 4: Student-work portfolios (Vermont)

Who takes them? Previously, students in grades 4 and 8 in math and English

What are they? Portfolios of student work

In the early 1990s, Vermont schools began collecting pieces of student work throughout the year. Students stuffed portfolios with letters they’d written, poems, and math problems, which were then sent to the state for review.

Sample questions:  In a fourth-grade writing portfolio, students had to include their best piece of writing, a letter explaining what they’d written, a written response to a book or current issue, and a poem, short story, or personal narrative, according to a 1992 report co-authored by Daniel Koretz, an education professor at Harvard.

What are the drawbacks? Hard to standardize

Vermont eventually scrapped the portfolio system. Officials decided that it was too hard to standardize them: The difficulty of the teacher-created math problems that students completed varied from school to school, for instance, and some students got help from their parents on certain projects while others didn’t, according to Koretz.

You can’t realistically compare a piece a student did alone,” he said, “with one that another student did with help from a parent — say, one with a degree in physics.”