Education Chat

Chat: How Can States Fairly Evaluate Teachers?

Under the Every Student Succeeds Act, states must decide the extent to which student test scores are included in teacher evaluations. Join us for a discussion on what states are doing and what the research says.

How Can States Fairly Evaluate Teachers?

Sponsored by:

Tuesday, February 7, 2017, 2 to 3 p.m. ET

Click here for more information about this chat.

Note: No special equipment other than Internet access is needed to participate in any of our text-based chats. Participants may begin submitting questions the morning of the chat.

Chat: How Can States Fairly Evaluate Teachers?(02/07/2017)

Close

2:59
Kim:
Thank you, Madeline -- and everyone for being part of today’s conversation. I wish we had time to take everyone’s questions.

Tuesday February 7, 2017 2:59 Kim

2:59
Madeline Will:
This was a fascinating discussion. Thank you so much for joining us, Kim!

Tuesday February 7, 2017 2:59 Madeline Will

2:58
Madeline Will:
And here’s Education Week’s recent article on how states are tinkering with their evaluation systems under ESSA: http://www.edweek.org/ew/ar...

Tuesday February 7, 2017 2:58 Madeline Will

2:56
Kim:
Here’s another great one:
Evaluating American’s Teachers (Popham)

Tuesday February 7, 2017 2:56 Kim

2:56
Kim:
Here’s my new book:
Student Growth Measures in Policy and Practice: Intended and Unintended Consequences of High Stakes Teacher Evaluation (Hewitt & Amrein-Beardsley)

Tuesday February 7, 2017 2:56 Kim

2:55
Kim:
10 Models of Teacher Evaluation (2014; Silverberg & Jungwirth)

Tuesday February 7, 2017 2:55 Kim

2:55
Kim:
Thank you, Kelly, for your interest in learning more. Here are some recommendations I have:
Getting Teacher Evaluation Right (2013; Linda Darling-Hammond)

Tuesday February 7, 2017 2:55 Kim

2:54
[Comment From KellyKelly: ]
Can you recommend any books or articles for further reading on the topic of teacher evaluation?

Tuesday February 7, 2017 2:54 Kelly

2:53
Madeline Will:
This is such a complex topic, and unfortunately we’re almost out of time. But here’s a good question from Kelly -

Tuesday February 7, 2017 2:53 Madeline Will

2:53
Kim:
Just the other day a district leader said to me, “You all have got to prepare more math teachers! We don’t have enough math teachers!” This climate of teacher shortages works against policy efforts to be tough on teachers through what are perceived to be more rigorous evaluation policies.

Tuesday February 7, 2017 2:53 Kim

2:51
Kim:
For example, in North Carolina we have seen a precipitous decline in student enrollment in teacher education programs. There were a number of schools across the state that still had teacher vacancies well into the beginning of the school year.

Tuesday February 7, 2017 2:51 Kim

2:51
Kim:
There are several reasons for my prediction: 1) Many teacher decry the use of student growth measures, and there is a general cultural sentiment in the US that undervalues teachers. These factors coupled with alarming teacher shortages in some areas (geographically and content-wise) make for a difficult labor market.

Tuesday February 7, 2017 2:51 Kim

2:49
Kim:
That said, I do not foresee a lot of states adding a student growth component to teacher evaluation if they don’t already have one, and I predict that states will -- if anything -- work to downplay or reduce the role of student growth measures in teacher evaluation.

Tuesday February 7, 2017 2:49 Kim

2:48
Kim:
I appreciate your question, Willie. I wish that I had a crystal ball. I predict that states will continue to rework teacher evaluation policy for the next several years at least, as they work to find a good solution to the challenge of accurately and fairly evaluating teachers.

Tuesday February 7, 2017 2:48 Kim

2:46
[Comment From WillieWillie: ]
Do you feel that many states will continue to allow teacher evaluations to be incorporate into students’ test scores due to the fact that there is no easy answer to this question.?

Tuesday February 7, 2017 2:46 Willie

2:45
Kim:
There is no requirement in the ESSA legislation about when or how states must determine teacher evaluation policy. In other words, it’s totally up to states now. States could further divest this role to districts, leaving it up to districts how they wish to structure their own teacher evaluation policy.

Tuesday February 7, 2017 2:45 Kim

2:44
Kim:
Hello, “E.” ESSA essentially mandated that the federal government could NOT dictate teacher evaluation policy (as it had done through Race to the Top, NCLB waivers, and School Improvement Grants). That role has been divested to the states.

Tuesday February 7, 2017 2:44 Kim

2:43
[Comment From EE: ]
When must each state make a determination on inclusion or to exclude student data for teacher evals?

Tuesday February 7, 2017 2:43 E

2:42
Madeline Will:
You mentioned that after the passage of ESSA, there has been a lot of movement in teacher evaluation policy, but states are all over the map. We’ve gotten a few questions about that:

Tuesday February 7, 2017 2:42 Madeline Will

2:41
Kim:
Teachers tend to like SLOs more than student growth measures that are based on standardized test because teachers typically have more influence on the measures and score goals for SLOs.

Tuesday February 7, 2017 2:41 Kim

2:40
Kim:
In some districts/states, teachers develop their own SLOs (both the measures/assessments they will use and the score goals) while in other places (e.g., Florida) there are certain tests that are approved by the state (e.g., for 4th grade social studies) that can be used.

Tuesday February 7, 2017 2:40 Kim

2:39
Kim:
Student Learning Objectives (SLOs) are notoriously unreliable and tend not to have reasonable evidence of validity. Also, there is a huge variety in the approach to SLOs.

Tuesday February 7, 2017 2:39 Kim

2:39
Kim:
These main approaches are 1) To use the aggregate school-level student growth score as the teacher growth score; from my research I learned that most teachers strongly dislike this approach, as they see it as invalid and unfair; 2) To use Student Learning Objectives (aka Student Growth Objectives, etc.).

Tuesday February 7, 2017 2:39 Kim

2:37
Kim:
I absolutely agree, Ryan! The majority of teachers do not teach in grades/subjects that are tested by standardized tests. There are a couple of main approaches that states and districts have taken to apply student data to these teachers’ evaluations.

Tuesday February 7, 2017 2:37 Kim

2:36
[Comment From RyanRyan: ]
It seems to me we’ve made little progress in finding the best way to include evidence of student achievement/growth in evaluations of teachers in non-tested grades and subjects. SLOs are notoriously unreliable and unproven in terms of validity, but they dominate the evaluations of most of these teachers (who make up about 70% of all educators). What do you think about this?

Tuesday February 7, 2017 2:36 Ryan

2:36
Madeline Will:
Here’s a question from Ryan about evaluations in non-tested grades and subjects:

Tuesday February 7, 2017 2:36 Madeline Will

2:34
Kim:
And while it makes sense that some teachers would be strong in some areas and not as strong in others, there tends to be little distinction in scores across groups of people. In other words, typically, teachers who are rated high in one standard/area, tend to be rated high in others. This isn’t universally the case, but it raises questions about how clear the distinctions of standards are in the minds of raters and how well-trained and skilled the raters (how tend to be principals and assistant principals) are.

Tuesday February 7, 2017 2:34 Kim

2:32
Kim:
You make a good point, Mark. New generation teacher evaluation systems typically include measures on various standards. In North Carolina, for example, there are five standards, including professionalism/reflection, pedagogy, etc.

Tuesday February 7, 2017 2:32 Kim

2:30
[Comment From MarkMark: ]
If the evaluation systems are oriented toward improving targeted areas of practice, not trying to give “holistic” ratings, shouldn’t their impact be more significant? No teacher is “effective” at everything - nor is a teacher “ineffective” at everything.

Tuesday February 7, 2017 2:30 Mark

2:30
Madeline Will:
Here’s an interesting observation from Mark:

Tuesday February 7, 2017 2:30 Madeline Will

2:29
Kim:
Teacher practice performance levels (typically based on observer scores and artifacts) tend to be more stable than performance levels based on student growth measures.

Tuesday February 7, 2017 2:29 Kim

2:28
Kim:
The research of Sean Corcoran and others, however, shows that teachers’ performance levels -- as indicated by student growth measures -- vary widely depending on the tests used as the basis for student scores and the methods (algorithms) used to calculate teacher effectiveness. Thus while there are more performance levels, these levels tend not to be stable across tests and student growth measures.

Tuesday February 7, 2017 2:28 Kim

2:26
Kim:
While historically evaluation systems have been dichotomous (effective or ineffective), the new generation of systems include more performance levels -- typically five.

Tuesday February 7, 2017 2:26 Kim

2:25
Kim:
I agree Ryan. Historically, less than 5% of teachers have been found ineffective on evaluations. Even under the new generation of evaluations, this figure remains under or around 5%.

Tuesday February 7, 2017 2:25 Kim

2:25
[Comment From RyanRyan: ]
One of the main drivers of changes to teacher evaluation was the perceived inability to differentiate quality with the old systems (see 2009’s Widget Effect paper for example). Yet in almost all new systems, the lack of differentiation is still present. Thoughts on this?

Tuesday February 7, 2017 2:25 Ryan

2:25
Madeline Will:
Ryan has a question about how to differentiate quality:

Tuesday February 7, 2017 2:25 Madeline Will

2:22
Kim:
It’s these very students and schools that need the very best teachers, and I fear that teacher evaluation systems that place strong emphasis on student growth measures will result in the exacerbation of the already inequitable distribution of quality teachers.

Tuesday February 7, 2017 2:22 Kim

2:21
Kim:
From my own research I have learned that the use of student growth measures in teacher evaluations tends to make some teachers avoid serving certain students (e.g., students with disabilities) and avoid teaching in certain schools (e.g., students with high percentages of students in poverty and students of color) because teachers feel that serving these students is risky for the teachers’ student growth scores.

Tuesday February 7, 2017 2:21 Kim

2:20
Kim:
Hello, Katie! I used to teach in Gwinnett County! From the emerging research we know that your concern about teaching to the test is valid. High stakes accountability systems tend to cause teachers to ramp up efforts to teach to the test.

Tuesday February 7, 2017 2:20 Kim

2:19
[Comment From Katie PardeeKatie Pardee: ]
With student test scores being factored into evaluations, how can we ensure that teachers do not “teach to the test”? The largest district in the state of Georgia (Gwinnett), announced that they would be transitioning to a teacher evaluation/compensation system that factors student growth. If test scores are tied to pay, what plans do schools across the nation have to attract teachers to low-performing schools?

Tuesday February 7, 2017 2:19 Katie Pardee

2:19
Madeline Will:
Here’s a good question about a potential concern with using student growth measures:

Tuesday February 7, 2017 2:19 Madeline Will

2:18
Kim:
There is also debate about the tests used to calculate student growth. Some tests do not have sufficient “spread” of questions (from below grade level to above grade level) to accurately identify a student’s performance, which makes it highly difficult to accurately measure a student’s growth over time.

Tuesday February 7, 2017 2:18 Kim

2:16
Kim:
There is a good deal of debate about which student growth measures -- which algorithms -- are most valid for isolating and identifying a teacher’s contribution to student learning.

Tuesday February 7, 2017 2:16 Kim

2:15
Kim:
Systems vary in what variables are controlled for in student growth measures. Some systems, for example, account for “student with disability” status or English Language Learner status. Others, like TVASS, do not account for such variables.

Tuesday February 7, 2017 2:15 Kim

2:14
Kim:
Hello, Willie! The two most common types of student growth measures that are based on standardized tests are value-added measures (e.g., EVAAS) and Student Growth Percentiles. In general, these systems compare student actual scores to either student projected scores or student past performance.

Tuesday February 7, 2017 2:14 Kim

2:13
[Comment From WillieWillie: ]
What indicators will make up or be considered as valid student growth measures?

Tuesday February 7, 2017 2:13 Willie

2:12
Madeline Will:
We’re getting some questions from the audience - keep them coming, folks! Kim, here’s one from Willie.

Tuesday February 7, 2017 2:12 Madeline Will

2:10
Kim:
Some studies have shown that teachers who have high percentages of English Language Learners, students with special needs, gifted students, and students multiple grade levels behind are more likely to have lower student growth scores.

Tuesday February 7, 2017 2:10 Kim

2:10
Kim:
Student growth measures are intended to be a measure of the teacher’s influence on student learning. That influence, however, is difficult to isolate.

Tuesday February 7, 2017 2:10 Kim

2:08
Kim:
While that argument is compelling, it is the mechanisms for measuring and incorporating student growth that have some experts calling foul. A number of prominent scholars and organizations (e.g., AERA, APA) have grave concerns about the validity and reliability of measures of student learning, such as value-added and Student Growth Percentiles.

Tuesday February 7, 2017 2:08 Kim

2:07
Kim:
There are cogent arguments on all sides of this debate. Some experts argue that the ultimate purpose of teachers is to growth students -- to help them learn. As such, it makes great sense from that perspective to include measures of student growth into educators’ evaluations.

Tuesday February 7, 2017 2:07 Kim

2:05
Madeline Will:
Interesting! Whether student growth measures should be included in teacher evaluations has been a controversial question. Can you explain that debate?

Tuesday February 7, 2017 2:05 Madeline Will

2:03

Kim:
Since the passage of ESSA, some state are postponing the inclusion of student growth measures (New York), some are reducing their percentage in the evaluation breakdown (Georia), or eliminating those measures altogether (North Carolina, Alaska).
Others, like, Tennessee is continuing with its teacher evaluation system, which incorporates TVAAS, a value-added system.

Tuesday February 7, 2017 2:03 Kim

2:02
Kim:
Since the passage of ESSA in late 2015, we are seeing a lot of movement in teacher evaluation policy, but there is no consistent pattern. Teacher evaluation policy is messy and complicated.

Tuesday February 7, 2017 2:02 Kim

2:02
Madeline Will:
The new federal K-12 law, the Every Student Succeeds Act, has removed federal requirements from teacher-evaluation systems, so states are free to decide for themselves how they evaluate their teachers. Kimberly, can you tell us a little about what that means? What have you been seeing from states?

Tuesday February 7, 2017 2:02 Madeline Will

2:01
Madeline Will:
Thanks for being here!

Tuesday February 7, 2017 2:01 Madeline Will

2:01

Kim:
Greetings, everyone! I’m honored to be with you today. I conduct research on the ethical and efficacious use of educational data. I’m particularly interested in how measure of student growth are used and misused.

Tuesday February 7, 2017 2:01 Kim

2:00
Madeline Will:
I’m joined today by Kimberly Kappler Hewitt, an assistant professor of educational leadership at the University of North Carolina at Greensboro. Kimberly, will you tell us a little about yourself and your research?

Tuesday February 7, 2017 2:00 Madeline Will

2:00
Madeline Will:
Hello everyone, welcome to our chat on teacher evaluations! My name is Madeline Will, and I’m the assistant editor for Education Week Teacher.

Tuesday February 7, 2017 2:00 Madeline Will

9:12
Mike Bock:
Good morning, and welcome to today’s free live chat, ‘How Can States Fairly Evaluate Teachers?’, sponsored by Measured Progress. I’ve just opened the chat for questions, so please start submitting them now. We’ll be back at 2 p.m. E.T. with Kimberly Kappler Hewitt and Madeline Will. Hope to see you then!

Tuesday February 7, 2017 9:12 Mike Bock

The Every Student Succeeds Act allows states to decide the extent to which teacher evaluations incorporate students’ test scores. But a year after the law’s passage, states are still tinkering with their evaluation systems to find the right balance. Student growth measures have become a controversial topic: Are they a way to hold teachers accountable to students’ academic progress based on metrics? Or are they unfairly punitive by reducing students to a test score?

Join Education Week for a conversation with Kimberly Kappler Hewitt, who researches teacher-evaluation systems and student-growth models. She will discuss how different states are choosing to grade their teachers and the research surrounding student growth measures.

Guest:
Kimberly Kappler Hewitt, assistant professor of educational leadership at the University of North Carolina at Greensboro

Moderator:
Madeline Will, assistant editor, Education Week Teacher

Related Article:
Assessing Quality of Teaching Staff Still Complex Despite ESSA’s Leeway

Share: