staying the course

Proposal to strip student growth data from teacher evaluations goes down

Republican Sens. Vicki Marble (left) and Laura Woods originally were co-sponsors of the proposed evaluation bill but had changed their minds when it came time to vote.

A bill that would have dramatically changed Colorado’s teacher evaluation system was defeated Thursday on a 6-3 vote by the Senate Education Committee.

Senate Bill 16-105, originally introduced with bipartisan sponsorship, would have allowed school districts to drop the use of student academic growth data in teacher evaluations. It also would have eliminated the annual evaluation requirement for effective and highly effective teachers.

Two Republicans who originally signed on to the bill voted no Thursday.

The requirement to base at least half of a teacher’s annual evaluation on student academic growth is a centerpiece of 2010’s landmark education evaluation law, Senate Bill 10-191.

That provision was hotly disputed then, and Thursday’s hearing demonstrated that passage of six years hasn’t fully cooled the passions.

Some lawmakers and education reform groups argue that use of student growth data gives a fuller, more objective picture of a teacher’s effectiveness than what’s provided only by a principal’s classroom observations and evaluation.

But many teachers, unions and lawmakers believe that use of student growth data is unfair, saying the tests used to generate that data are flawed and provide an incomplete picture. Critics also argue the evaluation law has placed a bureaucratic burden on districts, particularly smaller ones.

“This is not a valid method to evaluate teachers,” the bill’s chief sponsor, Sen. Mike Merrifield, D-Colorado Springs, told his fellow committee members. “We are not doing away with teacher evaluation. We are trying to change it so it is more fair and useful.”

The four hours of testimony and committee discussion resurrected arguments, beliefs and fears raised by the intense debate over SB 10-191 six years ago.

Sen. Mike Johnston, D-Denver
Sen. Mike Johnston, D-Denver

Committee member Sen. Mike Johnston, D-Denver, was the primary author of that bill. Six years ago, as chair of the House Education Committee, SB 16-105 champion Merrifield fought a losing battle against SB 10-191.

“This has been a great conversation. We had this debate before, but each year it gets more respectful,” Johnston said to Merrifield at one point. “Some things have stayed the same since this conversation started, and some things have changed.”

On Thursday, Johnston joined the five Republican members of Senate Education in voting against Merrifield’s bill. Two of those Republicans, conservative Sens. Vicki Marble of Fort Collins and Laura Woods of Thornton, originally cosponsored SB 16-105.

In closing remarks Thursday both said they’d decided the bill wasn’t the right solution, although they didn’t fully articulate why they changed their minds. Some conservative interest groups like the Independence Institute opposed the bill.

Two committee Democrats — Sens. Andy Kerr of Lakewood and Nancy Todd of Aurora — voted for Merrifield’s bill. Both were in the House six years ago and voted against the original evaluation law.

Merrifield’s attempt to rein in that law this year drew close attention from interest groups. It was supported by teachers unions and the Colorado Association of School Executives, which backed SB 10-191. Merrifield’s bill was opposed by multiple business and education reform groups, and many of their representatives testified Thursday.

Evaluation law has rolled out in slow motion

The evaluation law has been put into effect in stages and isn’t yet fully implemented, partly because of the complexities of setting up the evaluation system and partly because the 2014 launch of new state tests created a gap in the state data needed to measure student academic growth.

The 2014 legislature gave school districts flexibility in using growth data for the 2014-15 school year. Districts could use 50 percent, 0 percent or anything in-between.

The 2015 legislature made a different tweak in the evaluation law. In the current school year, districts are required to base 50 percent of evaluations on student growth. But last year’s testing reform law barred districts from using state testing data to measure growth.

That testing law also says that if school districts don’t receive state test results in time to use them for evaluations districts should use local measures of growth.

It’s a common misconception that student growth is based only on data derived from state test results. The original evaluation law required that growth be determined by “multiple measures” such as state tests, local tests and other data. The law also gives districts flexibility in how they weight the different data used to make up the 50 percent. Some districts use school accreditation ratings as part of the growth measure and apply them to all teachers in a school.

The evaluation system wouldn’t work without local measures of growth. Statewide tests are given in language arts and math, but only in grades 3-9. State science and social studies tests are given only once in elementary, middle and high school. And the majority of teachers don’t teach those subjects.

Three other evaluation-related bills are pending this session:

  • House Bill 16-1016 – Provides state help to districts to develop additional measures of student growth
  • House Bill 16-1121 – Exempts nationally board certified teachers from the requirement for annual evaluations
  • House Bill 16-1099 – Repeals a provision that requires mutual consent of a teacher and a principal for placement in a school a creates additional protections for teachers who aren’t place

There’s been statehouse chatter about extending the current time-out on use of state test data in evaluations, but no concrete proposals have surfaced.

Teaching teachers

Mentors matter: Good teaching really can be passed down to student teachers, new research finds

PHOTO: Jessica Glazer
Janet Lo (left) and Stacey Gong-Zhang attend a training program for pre-K teachers.

Do student teachers learn more when they’re mentored by especially effective teachers?

The answer may seem obvious, but there’s been little research confirming as much. Until now.

Three studies released this year offer real evidence that good teaching can be passed down, in a sense, from mentor teacher to student teacher. In several cases, they find that the performance of the student teachers once they have their own full-time classrooms corresponds to the quality of the teacher they trained under.    

And as many teacher preparation programs face pressure to improve, the findings offer a common-sense prescription: invest in finding the most effective possible teachers to supervise their trainees.

“Taken together, the point is that teachers who are … effective appear to be very promising mentors,” said Matt Ronfeldt, a University of Michigan professor who co-authored all three papers.

One of the studies, published last month in the peer-reviewed journal Educational Researcher, examined thousands of student teachers between 2010 and 2015 who were subsequently hired by a Tennessee public school. (Getting the data to understand this was a multi-year undertaking, since there isn’t a centralized system connecting mentors with their student teachers.)  

It found that teachers tended to be better at raising students’ test scores if their supervising teacher was better than average, too. Similarly, new teachers scored better on classroom observation rubrics when they had been mentored by a teacher who also scored well on that same rubric.

There was no evidence that teachers with more years of experience, all else equal, were more effective as supervisors.

The researchers can’t definitively prove cause and effect, but the results suggest that the mentor teachers are imparting certain specific skills to their student teachers.

The effect was small, though: Having a supervising teacher who did particularly well on their observations or their test scores was comparable to about half the performance leap teachers make between their first and second years in the classroom. That’s not a huge difference, but research has found teachers make their steepest improvement in those years.

A similar study, released in January, focused on about 300 student teachers in Chicago Public Schools who were subsequently hired in the district. Again, the student teachers who had better mentor teachers, as measured by classroom observations, ended up with better observation scores themselves.

Here too, there was no clear benefit of having a more experienced supervisor.

A separate paper, published in April through the research organization CALDER, looked at a single teacher prep program, Tennessee Tech University, which allowed researchers to conduct an experiment with its student teacher placements.

After all of the supervising teachers and schools had been selected, researchers divided them into two categories: those likely to be effective mentors and those less likely to be. This was based on data on the teachers (their performance and years of experience) and the schools (staff retention numbers and student achievement growth). From there, the nearly 200 teachers were randomly assigned, allowing the researchers to conclusively determine whether being in that high-quality group mattered.

It did. The student teachers with better placements reported that their mentor teachers were better instructors, offered more frequent and better coaching, and provided more opportunities for them to practice. This analysis didn’t track the student teachers’ later performance, but they did report that they felt more prepared to teach themselves and to manage their future classrooms.

This study, the researchers conclude, “would make a strong case to school systems that the quality of placements is fundamental to the development of new teachers.”

The set of studies add to a small but growing body of research on the best ways to set teachers up for success. Previous research had linked higher-functioning placement schools to better results for student teachers. Teachers also seem to do better after having student taught at a school with similar demographics as the school where they go on to teach. And concerns that adding a student teacher to a classroom hurts students (by allowing an untrained teacher to take over for a high-performing one) seem mostly unfounded.

The latest findings aren’t especially surprising, but to Ronfeldt they’re still important.

“While that may be a ‘duh’ moment, the reality is that there [are] often assumptions like this in education, and I think having the research evidence to back it up is critical,” he said, pointing out that few states have requirements that mentor teachers have strong evaluation scores. “We can make all sorts of assumptions, as I have for other things, and find out the opposite.”

Want to read more about efforts to improve teacher preparation? See Chalkbeat stories on teacher residencies, a Texas program known as UTeach, the challenges of identifying successful programs, a teacher training program that has embraced “personalized learning,” Denver’s effort to ease the transition into the classroom, and New York City and Memphis programs to recruit more men of color into teaching.

Tennessee Votes 2018

Early voting begins Friday in Tennessee. Here’s where your candidates stand on education.

PHOTO: Creative Commons

Tennesseans begin voting on Friday in dozens of crucial elections that will culminate on Aug. 2.

Democrats and Republicans will decide who will be their party’s gubernatorial nominee. Those two individuals will face off in November to replace outgoing Republican Gov. Bill Haslam. Tennessee’s next governor will significantly shape public education, and voters have told pollsters that they are looking for an education-minded leader to follow Haslam.

In Memphis, voters will have a chance to influence schools in two elections, one for school board and the other for county commission, the top local funder for schools, which holds the purse strings for schools.

To help you make more informed decisions, Chalkbeat asked candidates in these four races critical questions about public education.

Here’s where Tennessee’s Democratic candidates for governor stand on education

Former Nashville Mayor Karl Dean and state Rep. Craig Fitzhugh of Ripley hope to become the state’s first Democratic governor in eight years.

Tennessee’s Republican candidates for governor answer the big questions on education

U.S. Rep. Diane Black, businessman Randy Boyd, Speaker of the House Beth Harwell, and businessman Bill Lee are campaigning to succeed fellow Republican Haslam as governor, but first they must defeat each other in the 2018 primary election.

Memphis school board candidates speak out on what they want to change

Fifteen people are vying for four seats on the Shelby County Schools board this year. That’s much higher stakes compared to two years ago when five seats were up for election with only one contested race.

Aspiring county leaders in charge of money for Memphis schools share their views

The Shelby County Board of Commissioners and county mayor are responsible for most school funding in Memphis. Chalkbeat sent a survey to candidates asking their thoughts on what that should look like.

Early voting runs Mondays through Saturdays until Saturday, July 28. Election Day is Thursday, Aug. 2.