News and analysis from The Center for Michigan • http://thecenterformichigan.net
©2014 Bridge Michigan. All Rights Reserved. • Join us online at http://bridgemi.com

Original article URL: http://bridgemi.com/2013/11/when-11-isnt-2/

Talent & education

Rich school. Poor school. Whose teachers are better?

Perhaps the most vexing question facing Michigan’s new evaluation system is how to compare teacher performance when students in one class may be far more advanced than students in another. So-called “value-added data” is intended solve that dilemma, by taking into account such factors as poverty and a student’s past performance to determine a teacher’s impact on student growth.

(click image below to advance through slides.)

The complexities of student test scores in teacher evaluations.

The complexities of student test scores in teacher evaluations.

At the end of the school year, the third-grade students in Red School read at a grade level of 4.2, compared to a grade level of 3.5 in Blue School.

At the end of the school year, the third-grade students in Red School read at a grade level of 4.2, compared to a grade level of 3.5 in Blue School.

So the third-grade teacher at Red School should be rated higher than the third-grade teacher at Blue School?

So the third-grade teacher at Red School should be rated higher than the third-grade teacher at Blue School?

Red School is in a wealthy suburb, while Blue School has a high percentage of low-income students. Children of high-income, college-educated parents do better in school. Should teachers’ performance be adjusted for the income of their students?

Red School is in a wealthy suburb, while Blue School has a high percentage of low-income students. Children of high-income, college-educated parents do better in school. Should teachers’ performance be adjusted for the income of their students?

Red third-graders were reading at a 3.4 grade level at the beginning of the school year, increasing to 4.2; Blue third-graders started the year reading at grade level 2.5 and increased to 3.5. Should teachers’ performance be adjusted for net growth?

Red third-graders were reading at a 3.4 grade level at the beginning of the school year, increasing to 4.2; Blue third-graders started the year reading at grade level 2.5 and increased to 3.5. Should teachers’ performance be adjusted for net growth?

The Red third grade has one learning-disabled student, while the Blue class has four. Should that be taken into account too?

The Red third grade has one learning-disabled student, while the Blue class has four. Should that be taken into account too?

All the Red students were in the class all year; Blue had six students move into the third-grade class in February. Should the scores of those transfer students be counted in Blue teacher’s performance evaluation? Should they be counted as a fraction of a student?

All the Red students were in the class all year; Blue had six students move into the third-grade class in February. Should the scores of those transfer students be counted in Blue teacher’s performance evaluation? Should they be counted as a fraction of a student?

“It’s absolutely imperative that this be done with finesse, because people’s livelihoods are at stake.”

Dan DeGrow, superintendent of St. Clair County Regional Education Service Agency and former Michigan Senate majority leader.

Next slide

7 comments from Bridge readers.Add mine!

  1. ***

    “Red School is in a wealthy suburb, while Blue School has a high percentage of low-income students. Children of high-income, college-educated parents do better in school. Should teachers’ performance be adjusted for the income of their students?”

    Most likely it should be but is going to be very controversial.

  2. John S;

    There are a variety of evaluation designs that might be used but all are quasi-experimental designs with varying kinds of threats to internal validity. Perhaps the easiest to use is the before-after design (Yt-1 X Yt) with a pretest and posttest, with student learning being equal to posttest – pretest scores. The main threats to internal validity of this design are maturation and history. A post-test only comparison group design can be used, but there’s likely to be selection bias (students in different classes differing at the outset). Various control variables can be introduced for community, school, and individual level characteristics. The strongest design is an experimental design where students are assigned randomly to teachers. This type of experiment could be conducted at the school level (i.e., if there are two or more teachers at a given grade level) and the comparative performance of teachers evaluated and compared at that level. Perhaps principals should try doing it.

  3. Donna

    The truth is – teacher pay should not be linked to test scores. There are many variables beyond the scope of this slide show.

  4. TD

    Articles like this are why Ron French is becoming one of my favorite journalists. The best part of the achievement portion of the new tenure laws is the conversations between admin and teachers about using data to drive instruction and/or modify instruction. The system has improved but the full effects won’t be felt until layoffs hit. Teachers let go because of scores alone will crumble all the good that has come as a result of the laws. Principals need to be wise and resourceful applying them.

  5. Duane

    “…vexing question facing Michigan’s new evaluation system is how to compare teacher performance when students in one class may be far more advanced than students in another.”

    Disappointed again, Mr, French just does try to understand the question. It isn’t about comparing between classes or between schools, it is about comparing them to success of the learning.

    Success not trying to normalize practices by adjusting to challenges each student has is the desired measure.

    Why do we invest in anything, why to we pay for service, we are looking for value. Money why does someone buy a smartphone rather than a prepaid low end phone, it is about the value the want. It doesn’t matter what community they are in.

    Education is about learning, whine about all the barriers the student has to overcome (being from a poor family, a single parent or no parent family, a family that has little education, a disruptived classroom, etc.) and focus on the education of the students. Who has asked why is the a desparity of performance in the same classroom, Mr. French never seems to asked that question. Why are there kids from poor schools that go on to academic, financial, and social success while classmates of their fail at all three? Mr, French has never asked that question.

    Some students have more and larger barrier to overcome then others, but they do, why is Mr. French interested in why? Maybe the answers to that question will help teachers succeed more they how their adjusted score for those barriers.

    The other question Mr. French might ask is why do some teachers succeed while so many disappoint? Why do teachers recieving the same pay have such a difference in preformance, why are there poor teachers with tenure and why are there successful teachers with tenure? Mr. French doesn’t seem to understand success, motivation, ownership. Mr. French like much of the politics of education is focus on winning politically and are unwilling to put the effort into learning what is success, what does motivate, how does ownership change results.

    All the effort focused on managing compensation based on evaluation is like the tail wagging the ‘dog’. The reality is the ‘dog’ wags the tail when it is satified. Fignure out what success is, figure out why people strive for success, figure out how to inspire people to take onwership for what they do then you will have the ‘dog’ wagging its tail and you can figure out how to sustain success. You will also find how competion builds even better and better results and how it adjust to changing needs and barriers.

  6. Dedra

    There is a large variance in the physical plant, teaching resources such as computers, and even basic teaching supplies in a poor school and a rich school. How can these be accounted for in teacher evaluation?

    1. Duane

      Dedra,

      It would be easy enough to create a weighting factor for such things as supplies. But why should it matter, unless the evaluation system is about schools and teachers competing/being compared to others across the state.

      If the evaluation was against a model of success then supplies and other supporting factors would be part of the model and part of the evaluation process.

Leave your comment...

Your email address will not be published.

Invest in non-partisan journalism.

Donate to The Center for Michigan.Find out why.