Monday, August 16, 2010

Cheat Sheet for Principals

This was reportedly distributed to principals for their reference when discussing teacher contract negotiations.



TALKING POINTS FOR SPS LEADERS ON LABOR NEGOTIATIONS
OVERVIEW COMMENTS
• The vision for Seattle Public Schools is clear: We want every school to be an excellent school where every student receives excellent instruction.

• Nothing matters more to student achievement than good teachers.

• Our teachers deserve a system that recognizes their excellence and supports their growth as professionals.

• That is why we are proposing a new system to support all teachers at all levels of proficiency.

• We are building on our efforts in recent years to improve instruction through high quality aligned materials at all schools, professional development, and real time assessments of student learning.

NEW SYSTEM OF TEACHER SUPPORT NEEDED
• The next critical piece for improving instruction is clear. The District must hold itself accountable for results in student learning. We can do this by focusing resources on what research and experience tell us matter the most: supporting excellent instruction.

• A collaborative process is being proposed for Seattle where labor and teachers and school administrators come together to create a support system featuring a fair, thoughtful and balanced set of measures that help all teachers become even stronger.

• This system includes elements such as more collaborative time together, meaningful evaluations, recognition of teacher excellence, and specific strategies to support struggling teachers

• Supporting teachers’ success is at the heart of our goal to ensure that all students succeeding, so we must give teachers the tools they need to achieve as professionals.

• This includes using a balanced set of measures to give teachers meaningful and reliable feedback about their performance.

• We propose evaluations with multiple measures for four main areas:

1. Instruction and professional practice

2. Individual student growth, with multiple measures including District-required
and Teacher-determined measures

3. Whole school growth

4. Stakeholder evaluations

IMPORTANCE OF STUDENT GROWTH DATA TO A STRONG TEACHER EVALUATION SYSTEM
• When teachers have better information about how each of their students is doing, they are better prepared to help each student succeed.

• Teachers know how powerful good information about student learning is to their ability to reflect on their teaching and adjust to meet students’ needs.

• We have the chance to build the fairest and most reliable system in the nation, one that includes:

1. Multiple measures

2. Two assessments each year

3. Two year rolling average of data

4. Accounting for student demographic factors as we measure student growth

• We understand that teachers may want to wait and see how the system works before they opt in. So we propose that it be VOLUNTARY for existing teachers so they can decide if and when they want their evaluations to be tied to student growth.

• Tools like this will help us create a system that supports teachers who can teach to the gap – NOT to the test.

NEW RESOURCES ARE BEING SOUGHT TO BETTER SUPPORT OUR TEACHERS
• Supporting our teachers takes resources, and the district has demonstrated in recent years that even in tough economic times, we are strategic about using our resources so that student achievement is at the heart of our decisions.

• We have made extremely tough decisions in recent years so that we can dedicate every available dollar to our teachers in our classrooms.

• We have more than lived up to our commitment from previous negotiations to bring our teachers’ salaries up to competitive ranges.

• From the 2004-05 school year to 2009-2010, Seattle teachers’ salaries increased by 28%. The CPI during this time was around 16%.

• Ensuring that our teachers’ salaries more than kept pace with inflation and compared favorably to the region was important to ensure that we can retain and attract the best talent possible.

• Now that our teachers earn a professional salary (on average, over $82,000 including benefits), we need to focus on a system to ensure they are supported with professional development and practices that promote improving instruction at all levels.

• We are seeking federal Teacher Improvement Funds and will ask local voters for funds to support teachers in the years to come with programs such as:

1. For the 2011-2012 school year: 5 new STAR mentors, 2 HR consulting teachers, 208 career ladder positions, 150 service teachers -- total of $2,091,161

2. For the 2012-2013 school year: 5 new STAR mentors, 2 HR consulting teachers, 348 career ladder positions (including 17 higher-cost master teachers), 150 service teachers -- total of $2,560,361

3. For the 2015-2016 school year, we add approximately $4 million, assuming a scale-up of master teachers so that the lowest-performing 34 schools have at least a half time master teacher and the bottom 15 have a full time position. Having master teachers in every school would cost approximately $4.8 million in 2015-2016. The reason we are taking time to scale up is that we are requiring prior service as a demo and mentor teacher before qualifying as a master teacher

• These are just some of the elements we propose to add to support our teachers so that our students get the high quality education they all deserve. In addition to applying for a TIF grant and looking to local voters for supplemental levy funds, we work to continually increase efficiencies in our operations and remain active at the state and federal level to adequately fund public education.

TIMELINE AND MORE INFORMATION
• We all care about a strong and timely start for the school year. And we want to ensure successful instruction for all of our students. Because we know that good teaching is the path to student success, the district is ready to hold itself accountable to create a system that supports teachers’ professional success.

• Negotiations will resume again next week.

• We are committed to a good faith process at the negotiation table, and we will share information with the community on our website for labor relations at http://www.seattleschools.org/area/laborrelations/index.dxml.

• This information can also be accessed from the District’s home page by clicking on Labor Relations.

GENERAL TIPS AND SUGGESTED RESPONSES
• Start with statements about what we as Seattle Public Schools value:

o We respect the expertise and dedication of our teachers who serve our students every day.

o We are committed to offering an excellent education for every child.

• NEVER make negative comments about parties to the negotiation – no matter what.

o There are high expectations for the negotiators who come to the table, and each negotiation has its own dynamics.

o We are serious about increasing the quality of instruction in every school, and we know that we need significant change in the way we do business in order to meet our mutual goals of seeing all students succeed.

• Tie the concerns expressed back to the work we are doing to create a stronger system. Convey that we are working together on our students’ behalf.

o I hear how much you care about issue X. We understand how important this is to boost student achievement, and that’s why a meaningful and fair system to support teachers will make a solution to issue X possible.

• Do not speculate about what is happening at the table or why. Do not suggest new proposals or compromises that could be made.

o I understand you are concerned about issue Y being successfully resolved.

o Out of respect for a good negotiation process, I can’t speculate about what is happening at the negotiation table.

o But I can tell you what I know about the substance of our proposals, and everyone can look at the content posted on the web.

• Let people know where they can get more information.

o We will share information about our proposals on our website as we progress.

VALUES STATEMENTS AND CORE POINTS ABOUT NEGOTIATIONS
• Excellent instruction for every student is our core mission.

o We are all committed to providing every child an excellent education.

o We know that the quality of the teacher is the most important factor for student achievement that we control.

• Supporting excellent teaching is the heart of student success

o Our teachers deserve a system that supports them as professionals, recognizes their excellence, and encourages collaboration to strengthen instruction.

o That is why we are working together with labor representatives to make sure that every teacher gets the tools, time, support, and opportunity needed to help students succeed.

o Our proposal is that we continue to work collaboratively with teachers to complete a fair, thoughtful, and balanced evaluation tool that supports all teachers:

• We will offer paid stipends to recognize excellence and promote collaboration,

• We will increase support for principals to develop specific strategies to help struggling teachers improve their instruction by providing them with an additional $500 so they can access one-on-one mentoring or coaching or targeted professional development to help them improve their teaching practice,

• We will offer professional development and significantly more collaboration time—nearly 30 hours spread in throughout the year.

BACKGROUND INFORMATION
• HIGHLIGHTS OF RESEARCH ON THE CENTRAL IMPORTANCE OF TEACHER QUALITY

• One study looked at 8 year olds who started off at the same level of academic achievement. After 3 years, the students who had been taught by high-performing teachers scored 50 percentile points higher than those who had been taught by less effective teachers. The beneficial effect of a high quality teacher on student learning was even stronger for students struggling to meet academic standards.

• A recent NY Times article reported on a Harvard University study that found that the effects of a strong kindergarten teacher had an impact on children that lasted beyond their academic career, influencing their job prospects as young adults and even their lifetime earnings.

• On average, students taught by the lowest performing 5% of teachers only learn between ½ to 2/3 of a year’s worth of material. This means that two consecutive years of being taught by a low performing teacher can put a student a full grade level behind.

NOTE: an annotated bibliography of relevant research will be posted on the Labor Relations website

42 comments:

Charlie Mas said...

"Nothing matters more to student achievement than good teachers."
This is simply false. The primary determinant of student achievement is the active involvement in the student's education by an adult in the student's home. Right away, their second statement is false.

"The District must hold itself accountable for results in student learning."
I don't see the District holding itself accountable. I see the District trying to hold teachers responsible. How are principals held accountable? How are executives in central administration held accountable? How is the superintendent held accountable? How is the Board held accountable?

"A collaborative process is being proposed for Seattle where labor and teachers and school administrators come together to create a support system featuring a fair, thoughtful and balanced set of measures that help all teachers become even stronger."
There WAS a collaborative process developed, but the Superintendent tossed it aside in favor of her unilaterally designed process, SERVE.

"Individual student growth, with multiple measures including District-required
and Teacher-determined measures
"
What are the multiple measures? The descriptions we have seen rely mostly on MAP scores - a very unreliable measure of teacher effectiveness. Given that MAP is so unreliable, why is it one of the multiple measures at all?

"Tools like this will help us create a system that supports teachers who can teach to the gap – NOT to the test."
What does that mean? Those words don't make sense to me.

"We have made extremely tough decisions in recent years so that we can dedicate every available dollar to our teachers in our classrooms."
Yet the District continues to piss away millions on ineffective coaches.

"We have more than lived up to our commitment from previous negotiations to bring our teachers’ salaries up to competitive ranges."
So what do you want, a lollipop?

"From the 2004-05 school year to 2009-2010, Seattle teachers’ salaries increased by 28%. The CPI during this time was around 16%.
"
Which only shows how ridiculously low they were before. What does this have to do with anything? Salary isn't the controversy? Why are you raising this as an issue?

How are the new coaches (Master Teachers) different from the current coaches? How is this something new? Is this more teaching the teachers instead of teaching the students?

"We are committed to a good faith process at the negotiation table"
That's new and refreshing!

"we know that we need significant change in the way we do business in order to meet our mutual goals of seeing all students succeed."
How will principals and central administrators significantly change the way they do business?

"Our proposal is that we continue to work collaboratively with teachers to complete a fair, thoughtful, and balanced evaluation tool that supports all teachers:"
If the District is working collaboratively with the teachers, then why do the teachers say that this proposal is being imposed on them instead of one that was developed collaboratively?

"We will increase support for principals to develop specific strategies to help struggling teachers improve their instruction by providing them with an additional $500 so they can access one-on-one mentoring or coaching or targeted professional development to help them improve their teaching practice."
$500!? What will that buy?

"This means that two consecutive years of being taught by a low performing teacher can put a student a full grade level behind."
But Seattle Public Schools will promote them anyway.

dan dempsey said...

What a pile of BS... anyone buying this?


"We have made extremely tough decisions in recent years so that we can dedicate every available dollar to our teachers in our classrooms. "

Check OSPI for budget items:
SPS is the second largest spending district in the State by percent of budget spent on central administration (9%).

The percent spent on teaching is at best middle of the pack.

HERE is my current draft of testimony for Wednesday night.

another mom said...

"A recent NY Times article reported on a Harvard University study that found that the effects of a strong kindergarten teacher had an impact on children that lasted beyond their academic career, influencing their job prospects as young adults and even their lifetime earnings."

Is this the study that has not yet been peer reviewed?

The rest of the stuff leads me to believe that they think the public is just stupid.

Melissa Westbrook said...

I know that Dr. G-J is not going to be at the table but she'll be there in spirit. I believe this may turn into a "who will blink first?" negotiation and I'm not sure Dr. G-J knows how to blink.

She comes from a state without union representation and so far, it seems to be an annoyance to her. I don't look for a lot of give on her part.

Josh Hayes said...

I haven't seen comment on this, but Dr. Enfield was on "The Conversation" on KUOW at the noon hour today (Monday) - although it wasn't so much a conversation as an invitation to spew the party line.

Oh, heavens, no, this isn't about using a standardized test to assess teachers; we'll be using LOTS of standardized tests to assess teachers. And, uh, some other stuff.

I'm going ahead and making plans to teach my kids myself this September, if this is what passes for honesty from SPS. I'm not a reflexive "support the union!" guy, but in this case, the patently obvious slime from SPS is pretty damn off-putting.

wseadawg said...

They don't know what the hell they're doing, and no matter because they'll all be long gone when the chickens come home to roost. So, what is their for MGJ & Co to blink for?

They do not care about the kids in the classrooms, nor their families. They only care about more $$$$$$ in their pockets when they shoot on up and outta here.

To the envious on this list who bash teachers and their union and want "pay for performance", I ask: Does your pay for performance scheme at the job you probably hate reward or punish you not for what you can sell to your client, but how well and successful your client is at using and mastering your product?

Be honest. You know the answer. So stop wasting our time with apples to oranges analogies and go back to school until you understand what is and is not analogous to teaching children.

As if anyone from Microsoft would lose their job or be demoted because a customer didn't learn MS Word well enough. Gimme a break.

wseadawg said...

And how come all these brilliant ideas are nothing more than retreads from New York, Chicago, and New Orleans - all Broad subsidiaries.

How come nobody at SPS can identify a problem, analyze it using critical thinking, and develop a solution organically. Instead we get the Fox News-like "Research Shows" a la "some people say..." And critical thinking? Forget it. Not gonna happen with this group.

gavroche said...

Let's all play the "issue x" game with the Principals' talking points!

This is my favorite talking point on this list:

• Tie the concerns expressed back to the work we are doing to create a stronger system. Convey that we are working together on our students’ behalf.

o I hear how much you care about issue X. We understand how important this is to boost student achievement, and that’s why a meaningful and fair system to support teachers will make a solution to issue X possible.


Let's see how it works:

My first "issue X" would be poverty and how that affects a child's performance in school.

To which the canned talking point answer would be:

Principal-Bot:
"o I hear how much you care about poverty. We understand how important this is to boost student achievement, and that’s why a meaningful and fair system to support teachers will make a solution to poverty possible."

Is that so?

Hmm, I'm afraid that wasn't convincing. Let's try another concern of mine: the SPS Central Administration's gross mismanagement of resources and funds.

Principal-Bot:
"o I hear how much you care about the SPS central administration's gross mismanagement of resources and funds. We understand how important this is to boost student achievement, and that’s why a meaningful and fair system to support teachers will make a solution to the SPS central administration's gross mismanagement of resources and funds possible."

That didn't really answer my question either. Let's try another. Here's my next "issue x": What about the damning state audit of the Seattle Public School District that shows the Superintendent and Board are ignoring the law and not doing their jobs correctly?

Principal-Bot:
"o I hear how much you care about the damning state audit of the Seattle Public School District that shows the Superintendent and Board are ignoring the law and not doing their jobs correctly. We understand how important this is to boost student achievement, and that’s why a meaningful and fair system to support teachers will make a solution to the damning state audit of the Seattle Public School District that shows the Superintendent and Board are ignoring the law and not doing their jobs correctly possible."

Hmm, still not very convinced by this answer. Okay, here's an easy one: My issue x is smaller class sizes. I want them.

Principal-Bot:
"o I hear how much you care about smaller class sizes We understand how important this is to boost student achievement, and that’s why a meaningful and fair system to support teachers will make a solution to smaller class sizes possible."

Wish I could believe them. That answer didn't really make sense, actually.


(continued)

gavroche said...

(continued)

How about this one: I am disturbed by Superintendent Goodloe-Johnson's obvious conflicts of interest as a member of the Board of Directors of NWEA which sells SPS the MAP test, a member of the Board of the Los-Angeles-based Broad Foundation that supports privatizing and corporatizing public schools primarily through charters.

Principal-Bot:
"o I hear how much you care about Superintendent Goodloe-Johnson's obvious conflicts of interest as a member of the Board of Directors of NWEA which sells SPS the MAP test, and a member of the Board of the Los-Angeles-based Broad Foundation that supports privatizing and corporatizing public schools primarily through charters. We understand how important this is to boost student achievement, and that’s why a meaningful and fair system to support teachers will make a solution to Superintendent Goodloe-Johnson's obvious conflicts of interest as a member of the Board of Directors of NWEA which sells SPS the MAP test, and a member of the Board of the Los-Angeles-based Broad Foundation that supports privatizing and corporatizing public schools primarily through charters. possible."

Really?

Okay last but not least: My final "issue x" for today: Why can't Seattle Public Schools' Central Office focus on honest dialog with parents and teachers, about genuine, proven ways to help our kids learn, instead of obsessing over talking points and p.r. spin control?

Principal-Bot:
"o I hear how much you care about why can't Seattle Public Schools' Central Office focus on honest dialog with parents and teachers, about genuine, proven ways to help our kids learn, instead of obsessing over talking points and p.r. spin control. We understand how important this is to boost student achievement, and that’s why a meaningful and fair system to support teachers will make a solution to why can't Seattle Public Schools' Central Office focus on honest dialog with parents and teachers, about genuine, proven ways to help our kids learn, instead of obsessing over talking points and p.r. spin control possible."

dan dempsey said...

From the LA Times comes grading the teacher 9 pages of it. HERE

dan dempsey said...

The absolutely ridiculous part of SERVE is that the Principals, coaches, mentors, etc. are going to teach teachers how to do a good job ..... using Everyday Math, Connected Math, and Discovering

Absolutely insane ... the District has no clue how children learn math in the classroom.

The plan is to flush $4 million a year guided by these incompetents.

To improve a system requires the intelligent application of relevant data. -- W. Edwards Deming

$4 million a year will not fix stupid decision-making.

This District prefers politically based centralized planning like that found in the old Soviet Union (note the MGJ 5-year plan) and King County enables that District preference with a court system similar to that in use in Mainland China.

State Auditor discovers Stupid but there is just no fixing this degree of stupidity.

Recall the "5"

dan dempsey said...

From LA Times page 2:

The Times used a statistical approach known as value-added analysis, which rates teachers based on their students' progress on standardized tests from year to year. Each student's performance is compared with his or her own in past years, which largely controls for outside influences often blamed for academic failure: poverty, prior learning and other factors.

Though controversial among teachers and others, the method has been increasingly embraced by education leaders and policymakers across the country, including the Obama administration.

kprugman said...

The value-added analysis is controversial. It attempts to align for things that are alike and then look for positive correlations.

The rationale is that some classrooms do not have the same characteristics as other classrooms.

One of the problems with this protocol has been verification of the data that was used for the analysis. Researchers can pick and choose what data fits their analysis and it can be published anonymously. So there is increased bias.

The second problem is researchers are controlling for fewer variables. This is to reduce the complexity of their analysis, but this too increases bias.

The third problem are the test subjects themselves. The control group at the beginning of the study - is not the same control group at the end of the study.

Researchers have a difficult time finding classrooms where significant results can be achieved. Teachers and students frequently change classrooms and schools before the study has had time to collect enough information. Once again, the studies introduce bias into their results. That's why the majority of studies fail during review.

Pull the model protocols and revise the analysis. Stop biasing the data.

Eric M said...

Rally at the School Board meeting 5:30 PM, John Stanford Center for Educational Excellence (or, as some of us call it, "The John")

Wednesday. Tell the School Board what you think.

Maybe even talk to a principal-bot.

Recall petitions for the School Broad (oops, I meant Board, just a typo) ready yet? This would be a good place to gather some signatures, I think.

zb said...

They messed up, right, they're supposed to say "Teachers are the most important school-related variable" (I think that's on the official list of talking points from the reform movement). But they changed it and now it's obviously wrong.

Of course, Teachers are only the most important school-related variable of the ones that have been ranked and studied (not very many). For example, has anyone even done VAM studies of principals? of students? For example, if one did a study of the presence of one disruptive student in a classroom, would their presence affect the scores of other students in the classroom? If it did, what then?

Maureen said...

(Mainly in response to the LA Times piece Dan posted:)

Ok, here’s the deal. I am a huge fan of data-based decision making. I really am. But I am also very aware that it can be easy to manipulate data at all sorts of levels. There is also the basic question of whether the data you happen to have actually measures what you really care about.

I do believe that teachers can make a huge difference in the lives of children. I do believe that teachers that I personally rate as “very good” would create more “value” in the test scores of more of their students than would teachers I personally rate as “very bad.” I think that would probably be true almost no matter which test you used (including the walk in to the class and see if the kids are engaged test) as long as the teachers didn’t know they were being evaluated based on their students’ test scores.

In part because it is easy to manipulate data, I absolutely totally resist the notion of using any test’s scores to determine whether a teacher is “good” or “bad.” I do because then scoring well on the test necessarily becomes the underlying motivator and it is much easier to manipulate kids’ performance on tests than it is to truly be a good teacher. (See Diane Ravitch) I do because I question how well test scores actually measure real student success. I do because I believe it is the principals’ job to truly evaluate the quality of their staff, as it is any manager’s, and if they can’t do that well then they should not have that job. I do because I value teachers as professionals and I think the damage we would do to teachers by judging them based on their students’ performance on standardized tests is not worth the incremental gain of making the principals’ jobs easier.

And all of this is independent of the fact that a teacher could be “good” for one child or year and “bad” for another child or year. Again, it is a principal’s job to work with the teacher and help them do their best. The only possible justification I can think of to include student test scores as part of teacher evaluations would be to protect good teachers from bad principals.

Dorothy Neville said...

Bravo, Maureen! Very well said!

Meg said...

Maureen- as nerd who is fond of making decision based on data, you said what I've been struggling to articulate. Thanks.

zb said...

"I do believe that teachers that I personally rate as “very good” would create more “value” in the test scores of more of their students than would teachers I personally rate as “very bad.”

This just hasn't been my experience. I would have said that I have seen a range of teachers all of whom seemed perfectly acceptable, but that some of them were better for my child in a particular year and particular time (in a fairly unpredictable way).

But, I also think it's just possible that I'm looking at a small section of the overall distribution of teaching ability. Maybe I haven't detected anything 'cause all the teachers I've seen would fall into Maureen's top 20% or 10th, or whatever.

I've found at least one cite that points out that analysis of low-stakes testing might not transfer to high stakes testing and one that compared cross-validity of two different tests. In the cross-validity study (two different FL tests, one high, one low stakes), about 1/2 the teachers who were in the top quintile on one test were also in the top quintile on the other test. Since both test & stakes were changed, and repeated measures were used, we can't tell which of the possibilities contribute to the relatively low cross-validity.

reader said...

This is simply false. The primary determinant of student achievement is the active involvement in the student's education by an adult in the student's home. Right away, their second statement is false.


That's a mindless comment. They're factoring out the home contribution and focusing on the achievement gap itself as part of the agreement. Given the same home factors, IE, demographics, for a group of students, what's the teacher contribution? I'm not saying I agree with using standardized tests, but you can clearly factor out the usual problems people are complaining about. It would be so much better to give these incentive awards on a school basis, if you're using test scores as a measure. EG. The whole staff, that changes the trajectory of a whole building's test performance... should be incented (assuming the demographics in the building remain the same).

seattle citizen said...

Reader, here's an interesting take on "incenting" which states that research shows incentives to actually be a detriment when cognitive activities are being done. For menial or physical tasks, incentives seem to help, but for thinking tasks they hinder. Or so claims Dan Pink.

http://www.youtube.com/watch?v=u6XAPnuFjJc

dan dempsey said...

I DO NOT GET IT.

If there are teachers that are seriously deficient, someone needs to hold the Principals accountable.

This is beginning to look like a $4 million per year program that is needed because of substandard administrators.

=========
Why are teachers being bashed?

Mary said...

Interesting video on why "value added teacher assessment" doesn't work.
http://www.youtube.com/watch?v=uONqxysWEk8

hschinske said...

Value-added assessment *in origin* was a really good idea. They're just doing crappy things with it.

Helen Schinske

zb said...

"you can clearly factor out the usual problems people are complaining about. "

Your really can't. You can try to factor them out, and that can improve your analysis. But what we need to know here is whether you can factor out those features well enough to be able to use the VAM scores in a diagnostic way, to tell the good teachers from the bad teachers.

Mind you that's a very different analysis from looking at whether VAM-based teacher measures are relevant for something else, because the effect of the noise on decision making is different. The most optimized measures yet available are still unreliable. The bottom line question is whether the unreliable measures will actually have any positive effects on student outcome. For example, the cite below states that if bonuses were given to the top 20% of teachers based on VAM, only a 1/3 of those teachers would receive bonuses the next year, based on the reliability of the scores.

The correlations between rankings of elementary school teachers in adjacent years range from 0.08 to 0.36. The min is a negligible correlation (0.08!)to a max of 0.36 (meaning only 12% of the variation in performance is explained by the teacher).

http://www.urban.org/publications/1001266.html

These measures are simply not good enough for diagnostic purposes. And that's without even considering all the potential problems: Maureen's concern about gaming of high stakes tests, the fact that VAM can only calculated for teachers who teach tested classes, the tests themselves and their correlation with what we want the kids to be learning. The list is endless.

I think VAM scores are a useful area for continued research. The way I'd use them is as a screening tool to try to figure out which teaching methods might potentially be useful. Then we'd gather independent lines of evidence for what constituents best teaching practiced validated by multiple measures.

Dorothy Neville said...

Speaking of cheating...

Charlie Mas said...

Reader, you may have understood that they were factoring out the home contribution, but that's not what they said. As a reader, you are much better than I at changing what the writer has written to what you believe the writer meant. You have shown this talent at other times as well. I lack that knack. I don't know what they mean; I only know what they say.

Further down in the document they do include the modifier to say that teacher quality is the primary school-based determinant in student achievement. But that's not what they wrote in the beginning.

By the way, there has been some attribution analysis done on the academic achievement gap. That analysis has shown that no less than two-thirds of it - and probably more - is attributable to home-based factors instead of school-based factors. So you cannot focus on the gap without focusing on home-based factors.

Sahila said...

Teachers say MGJ pulled a swiftie on them... see here for details:

http://seattleea.blogspot.com/2010/08/are-you-communicating-to-your-school.html

reader said...

Charlie, it's useful to draw conclusions from writing. In fact, it isn't a "knack", it's what educated people do. I heard it that they were factoring in demographics out of the compensation plan from the district representative discussing it on KUOW. Will it really be in the contract? How could either of us know.

So what if 2/3 of the academic achievement gap is due to home factors. How not? Really it's probably more like 90%. That doesn't mean the school can't address the problem. It simply means that they haven't. Duh. What could possibly be more obvious. The point of public education is to provide an equalizer, and to provide access to the benefits of society. If we simply want to keep the rich, rich... let's get rid of public education all together and save the money. The private sector can already do that.

Dorothy Neville said...

"I heard it that they were factoring in demographics out of the compensation plan from the district representative discussing it on KUOW. Will it really be in the contract? How could either of us know."

See the Five Fast Facts about SERVE on the District website:

We will work with teachers to develop measures that account for the differences among students that can affect their academic
performance, such as family income, English language learning status, and special needs.


If I were a teacher, I sure would like this fleshed out and in writing before signing the contract, instead of just a promise that they will collaborate later to include it. Is that going to happen before the contract is finalized? Only the collective bargainers know for sure.

Charlie Mas said...

So now we have the idea that the student growth measure will be a measure of growth - did the student advance 1.0 years academically over the prior year.

Now we are told that the growth measure will be considered in the context of the student's demographic group (race, income, program), so that 1.0 years of progress might be the standard for some, 0.8 years of growth the standard for others, and 1.2 years of growth the standard for yet another student.

That's a lot of tricky calculations. Does the District have the resources to make these calculations?

And how will these various demographic standards for growth be set? From historical data? Would it be like: "Students in group A/3/0 have historically advanced .9 years academically for each year of school and therefore that is the standard for them." Or will it be grade specific? Will the District say that students in group A/3/0 have historically advanced an average of 0.9 years academically in the fourth grade and therefore that is the standard for those students in that grade?

Not only do I doubt that the District can do these calculations, I seriously doubt that they have the data to develop these measures.

Wouldn't such standards just perpetuate the historical results? Would it put teachers in competition with their peers to beat the average? Is there some objectively determinable standard that the District wants to impose? Does the District want to plant a flag and say that for students in group A/3/0 only an advance of 1.1 academic years will be acceptable?

There's a lot of confusing talk around this and precious little information. The District has a long and busy history of providing lip service to ideas and making promises of future action and then never actually doing any of the promised work. I would want these questions resolved before I agreed to any of this sort of thing if I were a teacher. I'd like to know which parts they really mean and which parts are just fluff.

reader said...

Right. It's a slippery slope. You've got a racial achievement gap. If you want to address it, you've got to acknowledge it. And that means... yes, acknowledging that some student groups perform LESS than a year's progress in a year. (otherwise, there would be no gap) When we acknowledge it, sometimes we also perpetuate it. And that's a risk. So, we've got the historical data... now, the proposal is to reward people who can change the trajectory of the historical data. Not a bad idea on the face of it... and, no easy task.

We've also got so-called advanced students, who are really simply "ahead". What's the point in having programs where the goal is to stay 1 year (or 2 years) ahead? When they are 35 years old, will it be usefull for them to say "we are working at a 37 year old level"? Acceleration means maintaining more than a year's growth in a year, year after year. To be eligible for something extra, the teachers need to maintain more than a year or 2's growth.

Figuring it out, is incredibly simple. For any group, figure out the trajectory... then figure out if the teacher beats the demographic she worked with. Voila. To my mind, the "figuring it out" isn't the problem. It's the teaching to the test that's a problem. But, isn't that already a problem?

I do greatly appreciate extra bennies going to teachers that take on challenges, especially at the school level. That seems to be a no-brainer part of the equation... and it would likely get a lot of these good teachers out of their cushy positions at high performing schools.

Maureen said...

I would pay money (some, not a lot) to hear Brad Bernatek address Charlie's questions.

reader says:Figuring it out, is incredibly simple. For any group, figure out the trajectory... then figure out if the teacher beats the demographic she worked with. Voila.

Please do this calculation for my daughter's 7th grade class. I have known most of those 60 kids since kindergarten and I think my general reflections would out perform any statistical analysis you could perform (even if you could get the data from SPS).

Demographic data is useful on a large scale but not on a human scale. True income (let alone wealth) data is not available. Mother's level of education is not available. Racial data can be meaningless. Health data is not available. Parents' marital (let alone mental) status is not available.

Je ne le vois pas là

Anonymous said...

It's not that simple. For the MAP test, the expected growth varies with the initial RIT score, or "initial status" of the student.

The 2008 NWEA MAP norm report gives the example that "students lower in achievement status tend to grow more than their peers who are higher in achievement status."
So the expected growth for a low performing student could be 20 points on the RIT scale, while for a high performing student it could be only 1 point.

Contrary to what one might think, the lower performing students should have the highest measured growth, while the highest performing students will most likely show the least growth. Does this mean that the higher performing students are not learning? Unlikely. It means they are already pushing the limits of the test. Their growth is not measurable in a significant way.

A "year's growth" could mean different things for different students. How will the District define a year's growth? Will it be based on the MAP calculation of expected growth for each student?

-anonymous reader

Charlie Mas said...

It's one thing for the District to say, while at the bargaining table, that they will set the growth expectations for each student based on the student's demographic (grade, race, income, and program), but that growth expectation should be bargained. The District should not have license to set it unilaterally or arbitrarily. It should be formulaic and it should be based on reliable data.

Only there is no reliable data.

Then again, the growth expectation can be readily calculated for each student as an individual by simply dividing the difference between the ending standard achievement for all students and the student's initial achievement status by the number of years to graduation. So for a fifth grade student with a starting achievement level of 4.4 and an ending goal of 10.0 (Washington state standards only run through grade 10, not grade 12), the difference is 5.6. With seven years to go until graduation, the target growth rate for that student is 0.8 years.

It works okay for students who are at or near standard in the early grades, but for high school students who are behind the growth target can be steep.

It also means that teachers with advanced classes are under almost no expectations at all since their students often achieve the 10.0 target in middle school.

Anonymous said...

Your analysis assumes a linear progression of growth over the years, yet measured growth is not linear. Once again, it's not that simple.

From 1st to 2nd grade, the median (50%) reading RIT score goes from 173 to 190, a 17 point difference. From 5th to 6th grade, the median RIT score goes from 212 to 216, only a 4 point difference. Over time, the expected growth per year generally declines.

A student's growth trajectory also depends in part on the initial rank of the student (which correlates to other factors). A student that has mastered much of the content being tested has less room for growth, while students with low initial scores have more room for growth. Their growth trajectories are not equal.

-anonymous reader

Charlie Mas said...

Anonymous Reader, yes. My point exactly. All of the District's fluffy talk about adjusting expecations for various factors is entirely meaningless without the details that describe exactly how these expectations will be set.

Only a fool would agree to contract with the critical details to be determined in future unilaterally by the other party. Surely the teachers are not such fools.

I'm thinking that this topic needs to be highlighted, expanded, and discussed separately in its own post.

Lori said...

Charlie, I'd love a thread to discuss the details (or lack thereof) of how these test-score-based teacher evaluations are going to work.

The SERVE announcement said it would be a 2-year rolling average. Does that mean you take the fall's average RIT score for the roughly 60 kids the elementary school teacher taught over 2 years and compare it to the spring average? If so, wouldn't large gains made by some students essentially hide lack of gains made by others? Is that "excellence for all"?

Or do you compare each student's fall score with his spring score and test whether there has been growth? Or maybe you use the students' percentile rankings instead of actual RIT scores. That's how the LA Times analysis was done. But is that appropriate? Is dropping from the 65th percentile to the 60th percentile significant? Is that the same as dropping from 95th to 90th?

They say they'll account for demographics and other factors that affect learning. How? Have they built the regression models that identify the impact of each important variable on outcome? If so, what population did they use to develop the model and then validate it? You'd have to use only students taught by an effective teacher to build the model, otherwise teacher effectiveness has to be a covariate, which gets us back to the whole cart-before-the-horse problem.

An absolutely fascinating study was commissioned by the Dept of Education (http://ies.ed.gov/ncee/pubs/20104004/pdf/20104004.pdf) that throws cold water on the accuracy of any of this. Schocet and Chiang assessed likely error rates in these data analyses and found that using 3 years of data for each teacher, the risk of making a Type I error with VAMs (that is, claiming a teacher to be ineffective when she is not) is over 25%. The Type II error rate (failing to identify an ineffective teacher) is of the same magnitude. You'd need over 10 years of data to have these error rates down to the 5% level that is used in clinical research. Is this okay with people? Is this paper even applicable to Seattle - because we don't know that the SERVE system would use the same mathematical approach to evaluating teachers that these researchers did.

All of these issues need to be resolved before anyone approves this contract. No one should vote on something that hasn't been adequately explained.

Lori said...
This comment has been removed by the author.
Lori said...
This comment has been removed by the author.
Dorothy Neville said...

Lori! Virtual Hugs from me!

I am a math person but only fluent enough in statistics to mistrust most naive statistical discussions. Your comments are the most clear I have seen and show a much better grasp of the statistical analysis than most.

Yes, I cannot imagine how the teachers could vote to accept this as proposed. How do you vote for a contract when important fundamental details are left as a "trust me, we will work it out later"?

That's also the biggest thing for me with considering capital levies. There is too much left for staff discretion with no trust left. The upcoming supplemental levy is for operations, but the more I look at it the less I trust it as well.

dan dempsey said...

""trust me, we will work it out later"

Same thing SEA Executive Director Bafia said about all the concessions made at Cleveland as well as the elements of Performance Management Policy approved on 3-17-2010.

Apparently SEA members enjoy paying $70+ per month to see their leadership roll over on most everything.