June 20, 2010

"The right way to assess teachers' performance"

Everybody now says we must evaluate teachers based on how much value they add to their students' performance, but nobody has actually done this. It can take a long time to work out a good statistical system -- e.g., it took Bill James 25 years to work out his Win Shares statistic for evaluating baseball players. Michelle Kerr, a teacher who thinks statistically, takes a first crack at some of the necessary conditions in this Washington Post op-ed.

Another tricky issue is what to do about teachers' aides, special ed aides, handicapped students' companions, and other adults who might be in the classroom with the teacher. The more adults the better (hopefully), but not all adults are of equal value to the teacher. Under a value-added system of pay, teachers will finagle even more than now to get the most useful aides in their classrooms. Those who win the political battles for the most and best aides will get paid more unless the aides' inputs can be taken into account.

35 comments:

l said...

Teachers' unions to America: The reason better people are not attracted to teaching is because the pay is low. Pay us more and you'll get better results.

Anonymous said...

It's insane to link employment and pay to testing when properties of the testing aren't yet understood, and no one understands how to measure value added. It would be like linking NBA salaries and roster spots to those models from economists you mentioned that had Rodman far above Jordan, and ranked Iverson as terrible player in 2000-2001.

Anthony said...

One problem with any sort of teacher testing is that students with different levels of intelligence will get different results, and in theory, this would lead to teachers trying to get the smart kids and dump the dumb ones, and that would lead to real race discrimination, among other problems.

Since IQ testing is politically incorrect, I propose that we use a proxy for intelligence when assessing the non-teacher inputs into a class: students' household income. It's pretty objective, it conforms to liberal shibboleths, and it actually has some correlation to student intelligence.

Anonymous said...

That was a surprisingly reasonable and constructive op-ed in the Post.

As for special ed teacher's aides, shadows, ad one on ones, schools generally try and distribute them instead of clumping them in one classroom, because aides for students with IEPs get paid out of district rather than school funds, so if the aide spends a lot of time helping out with the class in general, it's essentially the school getting an extra paid adult for free.

Not all aides are created equal. You get 10 dollar an hour minimally educated babysitters and sharp, highly educated specialists. And the better teachers usually end up with the better aides, for obvious reasons. At my kid's grade level we have a highly educated young woman acting as a shadow to an academically brilliant but socially stunted autistic kid and another, less educated aide helping out a learning disabled child who looks like a poster child for fetal alcohol syndrome. Guess which pair the good teacher called dibs on?

Anonymous said...

Those who win the political battles for the most and best aides will get paid more unless the aides' inputs can be taken into account.

And this differs how from any other bureaucracy [public or private/corporate]?

[The most successful bureaucrats (public and private/corporate) being the ones who are capable of assembling the largest and most powerful fiefdoms.]

But this is so depressing to think about - that 50 or 60 years ago, a single teacher could teach a room full of white kids [average IQ 100], without the need for any aides, or any school nurses administering Ritalin, or any adjunct liasons from the police department*, roaming the halls in an attempt to manage a school full of kids whose average IQ is somewhere in the range of 77.5 to 80, and trying to keep them from murdering one another.

Oh well, this nightmare isn't sustainable fiscally - the whole house of cards is about to collapse - and worrying about it too much is kinda silly and counterproductive.



*If you're new to iSteve, then you have got to see some of the cynicism at Second City Cop; for instance, this thread is a classic, and really restored my faith in the good old-fashioned common sense and decency and straightforward honesty of the Men in Blue.

BTW, Second City Cop is running a new piece** about how "if Chicago's West and South sides were their own cities, they'd be the deadliest and most violent in America".

Barack Hussein Obama's old stomping grounds, of Altgeld Gardens, is in Beat 533, way down in the bottom of District 5, and, oddly enough, is relatively safe, with only about 20 to 29.99 murders per 100,000 citizens per year.

Sorry, but I don't have any information about how William Ayers & Obama dished out the Chicago Annenberg loot by police district.



**E.g. I'm a big advocate for giant walls. I've been really practicing my German accent. "Vhere ahrr your papers?" I'm not kidding. Big walls. Certain people can't live in peace & coupled with failed policy we cannot allow this cancer to spread any more. Though what am I kidding, the sun almost set on our American Empire. My West Side district looks like a 3rd world country, except the kids are REALLY fat. Take that Sally Sturthers! In lock-up the other day I had to search under a guy's man tits. Totally gross. Ewwwwww!

Anonymous said...

This page at the CPD website indicates that Altgeld Garden's Beat 533 was tied for 62nd place, with only 2 murders in the last 12 months.

Similarly, the "Every Block" website shows three murders in 2008, two murders in 2009, and only one murder so far in 2010.

BTW, has anyone ever been arrested for either the murder of the Trinity United Church of Christ's choir director, Donald Young, or for the murder of Lt Quarles Harris Jr, who was cooperating with the feds in the investigation of John Brennan's infiltration of the State Department passport files?

Just curious.

Anonymous said...

Michele Kerr: Students who don't achieve "basic" proficiency in a state test be prohibited from moving forward to the next class in the progression. Students who can't prove they know algebra can't take geometry. If they can't read at a ninth-grade level, they can't take sophomore English -- or, for that matter, sophomore-level history or science, which presumes sophomore-level reading ability...

I suspect that my conditions will go nowhere, precisely because they are reasonable...


Michele, honey, your conditions will go nowhere because they would result in only low single digits' worth of Blacks and Mexican illegal anchor babies making being passed on to late-middle school and early-high school.

A student probably needs an IQ of about 110 to have any real hope of succeeeding at Algebra and Geometry [much less Algebra II and Trigonometry, which would definitely require an IQ out towards 115 to 120].

Best-Case Scenario

Average IQ: 80
Standard Deviation: 15
Required IQ: 110
Percentage Meeting Requirement: 2.275%

Worst-Case Scenario

Average IQ: 77.5
Standard Deviation: 12
Required IQ: 115
Percentage Meeting Requirement: 0.088903%, or about 9 students per 10,000.

Udolpho.com said...

Teacher performance ratings are a terrible idea for these reasons. In general, ranking systems don't work out and get relentlessly gamed and lead to far more politics. It takes years to get them to a basic level of reliability and sports stats have the fewest obstacles (people manipulating the assessments for ideological reasons, dumb educational fads, etc).

Performance ratings are a cop-out because people don't want to take on the unions and the administrations. They think performance ratings will be an end run around the two dysfunctional parts of education.

Anonymous said...

It's good that the author is poking holes in all of the various soft-headed ("Close the gap"!) conventional wisdom about education reform, teacher evaluations, etc.

But the Ravitch, Kerr, Sailer, etc., critique of the Bill Gates-Arne Duncan camp consistently underplays the utter intransigence of the establishment that makes firing even a catastrophically bad teacher impossible.

Anonymous said...

I'm afraid you and most others assume that all teachers are in fact trying to improve the education of the students. That is far from being true. In fact we pay a whole class of teachers to ruin the learning process.

They are called special education teachers. A few years ago I was dating a special ed teacher. Her job as she explained it to me was to force some damaged kid into a classroom where he/she was actively resisted by the school, principal, and teacher. The school recognized that these damaged kids would severely damage the learning environment for all the regular students. She not only didn't care she exulted in crippling a classroom of normals. That was her job as she saw it.

I was shocked but I think she was not in any way unusual. My understanding is that the whole special ed advocate movement assumes that they have to promote their kids at the expense of the undamaged.

Even if all this is true I have no idea how much of the problem it is. However I wouldn't be surprised if a large part of our educational shortfall was caused by the clashing agendas of various factions fighting it out in all those bloated school administrations.

Albertosaurus

headache said...

The obsession with teachers is nauseating. Basically I managed to dodge enough teachers in school to make it to uni undamaged enough so that I could start performing there. School was such a waste of time.

Steve Setzer said...

A lot of work has been done in this "value add" area already, Steve. You should use your favorite search engine to look for "growth model" (especially "Colorado Growth Model") and "William Sanders value added" -- Sanders has been analyzing "value add" for the Tennessee Department of Education since the early 1990s.

TGGP said...

Robyn Dawes thought up a method for this back in the 80s.

Anonymous said...

Duh.
Schools already know how to spot good teachers.
They just can't say it in public. They already measure teacher performance. That is why good teachers can so quickly get out of crap assignments and into better schools with better students. There is not salary incentive in public school. The only thing schools have to offer are their location and their students.

Good students attract good teachers.

Thorough study in the Journal of Labor Economics:

http://works.bepress.com/cgi/viewcontent.cgi?article=1010&context=c_kirabo_jackson

Anonymous said...

Boy you won't let up will you! Not everything is a proxy for race. This is about teacher performance.

Anonymous said...

I hope I don't come off as too mean spirited but I don't think it's very important to measure the quality of teachers. I do think teachers should be paid less. Oh dear, is that mean spirited too?

Like most people some time in their life I have done commissioned sales. This is true performance based pay and it doesn't require any fancy Sabremetrics type stats. I made more money than I ever had before working commission sales, but one day at 9:00AM I was expecting a $30,000 commission but it evaporated by 11:00AM. I never quite recovered from that shock. I guess I'm just not tough enough.

I don't think many teachers want that kind of performance judgement either. They and their unions will accept a bonus of course but not a fine - Johnny can't read? Fire his teacher. How do you like that?

I have supervised dozens of different job categories and most of them were not appropriate for any kind of tricky job assessment. I did hand out fairly substantial bonuses to programmers I directly supervised, but I had had hundreds of clerks reporting to me and I never had much idea of their performance. Clerks are considered to be doing OK if they do their jobs and don't cause trouble.

So, are teachers, especially grade school teachers, more like programmers or are they more like clerks? I would say clerks.

The whole idea that teachers can make a difference is a Hollywood plot line. As you have pointed out elsewhere, there is not much evidence for the existence of "star" teachers. Star teachers is such a silly notion that only a culture that is desperate to find some cure for poor NAM school performance would ever entertain such an idea. Is the school administration too clumsy and inefficient? Let's find some star clerks.

Public school teachers don't choose their own texts or design their curriculum. Unlike a programmer they do not start with a clean sheet of paper and must produce something out of their own resources. For example, a year or so ago I read a sympathetic biography of Aaron Burr. Could a teacher assign this book and not get caned? Hardly?

Clerks have only a few "degrees of freedom" - so do public school teachers. At the college and university level, things are different but public grade and high school teachers are supposed to be factory workers. If a teacher had some really great creative ideas - like let's go picket Sarah Palin - he would be on Bill O'Reilly licketysplit.

Unions are institutions designed to raise wages above their free market values. Teachers make too much money and class sizes are too small. Spend less and forget teacher performance measurement.

Albertosaurus

Anonymous said...

Albertosaurus - Clerks are considered to be doing OK if they do their jobs and don't cause trouble.

I think that applies to a lot more than just clerks!

Keep your head down, do a competent job, show up every day, that sort of thing.

Anonymous said...

Unions are institutions designed to raise wages above their free market values.

At some meta level are not all institutions, unions, economic actors etc designed to raise their returns above free market values? In a sense thats what the free market is, the outcome of these interactions.

Robert said...

I have seen data that showed that the more female teachers a nation had the worse that nation's results in mathematics.

John Seiler said...

The real problem is schools themselves. Get rid of them, as John Taylor Gatto says:
http://www.lewrockwell.com/north/north853.html

OneSTDV said...

Basically I managed to dodge enough teachers in school to make it to uni undamaged enough so that I could start performing there. School was such a waste of time.

Agreed. I learned nothing until I got to college.

OneSTDV said...

I think the testing issue is pretty simple: Give kids the tests one year as a baseline. Then give out pay based on RELATIVE performance. And while this is premised on HBD and intelligence differences, it doesn't necessarily need to be supported on those grounds.

ranked Iverson as terrible player in 2000-2001.

Iverson was terrible that year and every other year for that matter.

In Bball, team success is generally a very good indicator of individual success. Look at what happened when Iverson got traded for Billups. Nuff said.

keypusher said...

You covered this already, Steve. There are 3.5 million teachers in America, which means the bulk of them are going to be utterly average. There are fewer than a thousand major league baseball players, all on the far right of the bell curve. That's what makes sabermetrics worthwhile.

Henry Canaday said...

How about we handle this the way we handle every other area where organizational performance is important, but individual performance is difficult to measure precisely:

We let parents choose the schools their children are to attend.

We let the organizations that run schools choose the principals.

We let principals choose how to hire, fire, promote and compensate teachers based on the principals’ judgments, informed by statistics and observation and subject to their contracts with teachers.

The government does what is does in other areas. It subsidizes the tuition of children whose parents are too poor to pay reasonable charges. It prevents fraud, ensuring that schools are delivering what they say they are delivering to parents. It prevents exclusion by race, but does not concern itself with racial balances in courses, classrooms or schools.

Poor parents would most likely seek schools that return to them at the end of the day a well-behaved child. Gee, come to think of it, that, along with basic literacy and numeracy, is probably the most important thing schools can do for poor children.

I think this is roughly the Murray-O'Rourke solution. When a sociologist and a clown agree, pay attention.

Gene Berman said...

Albertosaurus:

I presume you meant "canned," rather than "Caned." Although, come to think on it a bit...
(the Singapore solution?)

Camlost said...

ranked Iverson as terrible player in 2000-2001.

Iverson was terrible that year and every other year for that matter.

In Bball, team success is generally a very good indicator of individual success. Look at what happened when Iverson got traded for Billups.


Exactly. And look at what happened to the 2004 Olympic team when they added both Allen Iverson and his doppleganger Stephon Marbury - they became the first "Dream Team" of NBA players to lose in the Olympics.

It's not good to have a small, underweight, shoot-first point guard who ties up the ball for 18 seconds of the shot clock before firing up a fade-away or floater.

The only reason Allen Iverson even made it to the finals once with the Sixers is because of the sheer coaching brilliance of Larry Brown, and NO other reason.

Gene Berman said...

"In a sense, that's what the free market is, the outcome of these interactions."

In a certain sense, you're right--that's how markets operate in the (only nominally free) markets with which we're familiar, which fall distinctly short of qualifying as trult "free."

The forces operating on a market produce "market prices" and include both buyers and sellers (and, to a degree, those who'd be buyers or sellers but are precluded by the height of prices.

For markets to work freely (and to thus make possible the benefits of free markets), participants must be afforded both those things we call "the rule of law," (protection of persons and property against fraud, theft, and threat of or actuality of violent aggression against either, including a system of enforcement and adjudication of disputing claims, etc.) AND "equal protection of the law," (without which the law itself is accesory to various infringement of the unhampered (i.e., free) market.

By any impartial view, whether of facts or theory (or combination),
unions are criminal actors in the marketplace, lately legitimized and even legally-advantaged by derelict civil authority. This is especially notable in the case of certain types of union. Even long after unions had been recognized as "legitimate," it remained felonious for policemen, firemen, teachers, etc. (generally the segment "public service employee" or "government employee") to form or join a union or to act in concert (conspire) against their employer (the rest of the p[ublic, as represented by a particular administration.). Today, that category includes most membership.

Unionization is nothing but gang violence and intimidation usually employing "business agents" (their name for enforcers) and lawyers both to present demands, artfully describe the consequences of noncompliance with demands, and work out the details of capitulation ("negotiations").

What union members get in the form of heightened wages (shared generously with leaders and lawyers--including aforementioned enforcers) comes only slightly (if at all) at the expense or employers (as a generality, though some may be crioppled or hurt severely) but mostly at the expense of consumers in general and especially to the detriment of those willing workers who might have replaced intransigent union workers if not for the projection of violence (or, for about 70+ years now, favoring by the law.)

In defense of unions, it was often advanced they acted to insure a level of competent workmanship--a benefit to society--similar to that exercised by certain trade or professional organizations and the guilds of yesteryear. Those claims are, in the main, true (but carried with them the "capture" of membership by families and ethnic groupings to exclude entry by some among "outsider' groups who might have, become qualified).

Gene Berman said...

John Seiler:

It isn't schools that are the problem, as you undoubtedly are aware (but didn't express).

The problem is tax support of both buyers and sellers and "rigging" the market by making "buying" compulsory.

Chicago resident said...

I have a novel idea: Why not just ask the students which teachers are the best? After all, they're the ones who would know. Most of these schemes to evaluate teachers never think to get any input from students beyond what they can do on a test. As a student I was aware of who was a waste of time and who was informative. I guess it's like dog training, nobody asks the dog if it feels it got something out of it. If it acts more obedient then the training is regarded as successful.

BamaGirl said...

"They are called special education teachers. A few years ago I was dating a special ed teacher. Her job as she explained it to me was to force some damaged kid into a classroom where he/she was actively resisted by the school, principal, and teacher. The school recognized that these damaged kids would severely damage the learning environment for all the regular students. She not only didn't care she exulted in crippling a classroom of normals. That was her job as she saw it.

I was shocked but I think she was not in any way unusual. My understanding is that the whole special ed advocate movement assumes that they have to promote their kids at the expense of the undamaged."



Correct. And sadly, this has been going on for at least 15 years in public schools (maybe longer). Until high school, I typically had to deal with the distraction of one or two special-ed individuals in my class each year. They would make noises, yell out inappropriately, and occasionally get violent and pitch fits. Their aides were typically fat rude black women who I recall forcing the special kids to the middle of the cafeteria tables during lunch despite the fact that it made everyone else extremely uncomfortable.

Florida resident said...

John Derbyshire wrote recently a good review of the book by Robert Weissberg "Bad Students, Not Bad Schools"; see Derb's web-site.

After reading it, I bought and read the book. I like the book.
My comment on Derbyshire's review.

Derb somewhat ignored the last chapters of the book, where the author (Weissberg) says,
that we are doomed to live in the given demographic situation, and we have to pay welfare (or make work) racket to teachers and other school-related personell, as not to get riots.

His (Weissberg's) main suggestion in this (doomsday) situation is not to reform existing school system for masses, since it is hopless to attempt, but to improve education of gifted, not dissolving it by diversity requirements.

I do reccommend the book.
Florida resident.

Unknown said...

Ah yes, the value added. My school is going "back on the list" for failure to reach value added expectations, which the more astute of us love, because we'll probably spend the next four years going up, which will look good in the papers. Then around year 4 or 5, we'll plateau, fall from grace, and the cycle repeats itself.

On SPED teachers, guidance counselors, assistants, title 1 directors, and other untested personnel, they are largely worthless. I've been at 4 schools in my career, and I've never met a one of these people who were worth a damn. Many of them avoid coming in the classroom, or pull out compliant youth and sit in the teachers lounge.
If you teach in a rural school, the tradition is that the English teachers wind up doing the jobs of most of these folks: after school tutoring for free, scholarship applications, grant writing, etc.

I quit being an English teacher and moved over to science and foreign language so I wouldn't have to do that anymore. It was the norm at the school where I grew up, as well as at the first two schools where I taught.

888 said...

I think Seiler and Canaday are on the right path. We should have a system that respects the liberty of parents, schools and teachers: parents to choose the schools, schools, the principals, principals the teachers, and teachers, the curriculum.

At the same time, parents who don't need/want a school to educate and babysit their children, must have the option to teach them at home, and to have them work some limited hours.

More to the point, if children were paid 75-100 dollars a week to prove proficiency in math, reading, history, sciences, and foreign language, most will be done with high school level by age 14 or 15, even while getting several hours of exercise (play) and work in each day.

Leer said...

Help me understand, are these "value-added assessements" still accurate if they don't take into account individual IQ and peer effects (basically how many bad students are in the classroom)?

Mike Courtman said...

Michele Kerr: Students who don't achieve "basic" proficiency in a state test be prohibited from moving forward to the next class in the progression. Students who can't prove they know algebra can't take geometry. If they can't read at a ninth-grade level, they can't take sophomore English -- or, for that matter, sophomore-level history or science, which presumes sophomore-level reading ability...

We kinda have this education system downunder, low-achieving students now do the same material over and over and with constant re-sits and they can't leave school under they are 17.

In more humane countries like Finland and Germany they let these students go to non-academic schools and learn to do something practical.