Teachers blast L.A. Times for releasing teacher effectiveness rankings

[maddezmom]

Source: LATimes

National and local teachers unions sharply criticized The Times on Sunday when the newspaper published a database of about 6,000 third- through fifth-grade city school teachers ranked by their effectiveness in raising student test scores.

"It is the height of journalistic irresponsibility to make public these deeply flawed judgments about a teacher's effectiveness," said a statement issued by United Teachers Los Angeles.

The database is part of a Times series that rated teachers by using a "value-added" analysis based on seven years of standardized test scores obtained from the Los Angeles Unified School District. The value-added method looks at previous student test performance and estimates how much a teacher added to or subtracted from a student's progress.

By late Sunday afternoon, the database had generated more than 230,000 page views, an indication of the interest in the issue because Web traffic tends to be higher during the week.





Read more: http://www.latimes.com/news/local/la-me-teacher-react-20100830,0,1507297.story?track=rss

[dkf]

1. In a way you dont need to release teacher scores because the kids scores say it all.

[madfloridian]

20. What do you think they are releasing???

The students' test scores are used, but only the teachers are blamed if they are low.

Think about that a minute.

[proud2BlibKansan]

46. Do you understand what they are releasing?

Interesting how so few seem to understand what's really going on here.

[Demeter]

2. Because Raising Tests Scores Is ALL that Education Is About

:sarcasm:

Only an ignorant idiot would think so, by the way. Someone like W, who designed this folly, this fallacious policy. And anyone who buys into it, whether for political reasons, or because they drank the Koolaid.

[backtomn]

19. I don't agree.

If a test includes the important information that a child SHOULD know, in that subject, at that year, then I don't see a problem. How we use the information is what is important. There are a lot of factors involved in the results, but it doesn't mean that we shouldn't try to figure them out. This is fairly important.

[LostinVA]

3. I still don't see how this doesn't break privacy laws

If this is related to performance issues. HR records aren't public.

[Jefferson23]

4. Propaganda, how thoughtful and obliging of you LA Times. n/t

[Mass]

5. It always amazes me to see how people can put so much weight on useless measures.

If you want to know if a school works well, check kids's scores from one year to the next from grade to grade. There are still problems with these measures, but at least they compare the same kids year after year and show what the school did to make them progress.

But comparing test scores from different students is absolute idiocy, aside from all the other issues.

[Lightning Count]

8. That's pretty much what they did.

The value-added method looks at previous student test performance and estimates how much a teacher added to or subtracted from a student's progress.

[backtomn]

17. Scores grade to grade don't help much....

Each grade level involves more difficult and complex learning. A child could do worse in 5th grade than 4th, but would merely have made no improvement (rather than an incorrect assumption that they were doing worse).....or may have reached a limit to their ability to learn as quickly. I am for all the testing that is possible, but how we use the results is the important thing.

[msongs]

6. kid test score = ONLY how kid chose to answer questions at THATmoment nt

[woo me with science]

9. Yes, and if the child actually KNOWS how to add,

he or she is more likely to "choose" the correct answer.

[Iwillnevergiveup]

7. My former 8th grade student,, Salvador, comes to mind

His father frequently kept him out of school to help him with his roofing work.

[mbperrin]

10. Nothing sillier than the value-added concept.

Any measurement means you have to control all factors except the variable you're measuring.

For instance, in Texas, the 10th grade TAKS scores count for the school rating, but count nothing for students. They know this, and scores tend to be rather ho-hum, indicating that sophomore teachers are not very good, yes?

Then the 11th grade TAKS scores don't count for the school rating at all, but the student must pass them all to graduate. Amazingly, those great 11th grade teachers get those scores up in an amazing way! Fantastic!

Never mind a school like the one I teach in with a 20% special education population. Value-added will show smaller progress than the AP/IB students, yet those students and teachers are doing all they can, and they are getting wonderful results year on year. But numbers show less performance gain than the high achievers, so their teachers must suck, right?

All simple-minded people want a magic number, but education is far more complex than that. You don't really know you did for 20+ years, when the kids you taught use what they learned (or not). But reality is a lost concept in education and the rating of teachers.

[Lightning Count]

12. The solution is probably to have all years test count toward the school rating. nt

[Doremus]

14. The solution is to throw the fucking tests out and teach like they used to.

You know, back when American schools were the best in the world.

Then.

[Lightning Count]

15. So no national standard whatsoever and no measure of effectiveness?

Does not seem like a good idea. By the way, the SAT has been around since 1901 to give some perspective on standardized testing. We need school more like Japan, Canada, and Sweden.

[boppers]

28. "back when American schools were the best in the world."

When, exactly, was that?

[mbperrin]

62. Perhaps when their students produced more goods than any other country,

created a demand for living here that is still not sated, walked on the moon, and other singular accomplishments.

Results? Those are real.

[mbperrin]

16. One test, one day. What happens when a child is feeling ill that day?

Mom and dad had a screaming match the night before the one test?

Electricity was cut off the week before the test?

Sewer pipe bursts at school and floods entire building 2" deep? (Actually happened here last year.)

85% of our students qualify for free or reduced lunch. Might the difference in input resources make a difference compared to my granddaughter's district of 300 kids in K-12 with an average income of $160,000?

The whole magic number concept is simply too flawed to be applied in other than the broadest strokes - it certainly cannot come down to individuals.

[Lightning Count]

18. Then it will average out.

Every school has these problems. Unfortunately, without some sort of national testing standards you will end up with widely disparate schools such as those in Westchester, NY vs. Memphis, Tennessee. There has to be some way of telling when an "A" really is an "A" when compared to students across the nation.

[mbperrin]

23. Every school does not have a 20% special ed population nor an 85%

free or reduced lunch population. Mine does. My county also leads the state in unmarried teen pregnancies and is second in the state in STD transmission to teens.

Why must there be one standard? Why treat students all alike? They're really all quite different, and no two educations yield the same results (and how boring it would be if they did.)

[Lightning Count]

24. Should make zero difference.

The idea is to establish what constitutes an "A" across the entire nation so there isn't such discrepancy between how schools grade. Why must there by one standard? Because otherwise you end up with a double standard.

There also must be a consistent measure of teacher performance. Schools in Canada, Great Britain, Japan, Sweden, etc. somehow make due. I'm sure we can too.

[Occulus]

30. .

oops

[mbperrin]

34. Never taken a statistics course of any kind, have you?

Do, and when you return, you can strike the "should make zero difference" flip remark and actually discuss the topic.

Again, what's wrong with multiple standards? Why must the future plumbers of the world take as much foreign language as the future language teachers of the world? Why must the future English teachers of the world take as much math as the future math teachers? They do in Texas, the home of one-size-fits-all education.

Why do we have a 30%+ dropout rate? Students and parents find the curriculum irrelevant, therefore boring and a waste of time, to their future plans.

Since you took on the burden of foreign evaluation systems, find where each of those countries ranks, and then look at how they evaluate teachers. Return and report. You put it out there - it's your burden.

Until then, have a nice day.

[Lightning Count]

36. How many college professors will take your test score under consideration if you:

Have a fight with your girlfriend?
Had to attend that cool keg party?
Watched an all-night marathon of Ren and Stimpy?

Same principals apply to HS. Regardless of the problems the night before. Schools across the nation have to be made to adhere to the same standards. You can't have the valedictorian of one school be equal to the 100th ranking of a similar school all based on how teachers want to pad the scores.

If you want to campaign to open teenage plumbing schools and teaching schools then go for it and those schools can adhere to the standards geared towards those schools. We are talking about general education though.

If students drop out then so be it. An "A" needs to be an "A" across the board.

Let's start with Sweden.

http://www.prim.su.se/artiklar/pdf/Sw_test_ICME.pdf

[mbperrin]

40. Thanks for writing off all the 16, 17, and 18 year olds in the country.

Your insistence that grades are more important than people fits perfectly with current thinking at the "top" of the education field. At the high I teach, and attended in the 60s, we had plumbing, electrical, welding, paint and body, printing, and more, all leading to certificates, ensuring they were ready to go to work in those fields. That was from 1946 until 1984, when the stupidest idea ever became Texas policy, and now national policy: everybody, regardless, needs to go to college.

By the time plumbers finish their 4,000 hours of practicum, why do they need 4 years of English, science (including physics), math (including calculus)? ALL students must take all those courses now in Texas as part of the 4x4 rigor and vigor system.

Call a plumber next weekend in case you think it's an unworthy profession. We had a root clog, cleared up in 30 minutes. Weekend - 4 hour minimum @ $90 per hour for a total bill of $360. (Regular rates are $60 per hour during the week, a little over $120,000 per year, among the top 10% of wage earners in the US.)

[Lightning Count]

45. So basically the USA knows best.

And you seem to be against well rounded students.

Unfortunately, Americans seem against taking any advice from anyone. Maybe if we did, we wouldn't be middle to low of the pack.

http://en.wikipedia.org/wiki/Programme_for_International_Student_Assessment

[mbperrin]

42. Sweden. Your link has all kinds of explanation for non-compulsory student

testing, a new grading system, and no indication that student test scores are used in the evaluation of the teacher at all.

Next.

[Lightning Count]

47. I have been arguing that we need grading standards in school.

As you can see from my previous posts.

Would you be supportive of any standard measure of school teachers?

Does this sound like your average day of teaching?

http://www.emu.dk/gym/fag/en/internationalt/englishinsweden.pdf

[mbperrin]

52. The thread is teacher evaluation based on a single student test.

That's what you're promoting. That's what I'm against.

[wtmusic]

55. No, that's not what it's about.

It's about using student test score improvement to evaluate teachers, and it's over a period of seven years. Many tests.

[wtmusic]

38. Very few primary- or secondary-level students even know what the future holds.

The future English teachers have to take as much math as future math teachers, because the math that's taught through secondary school is used by everyone. It's part of a well-rounded education.

"Students and parents find the curriculum irrelevant." That's why we have education professionals making up the curriculum - not students and parents.

[mbperrin]

41. Did you use your calculus today?

Or your knowledge of King Lear?

Just part of the required toolbox in Texas for all students these days.

[rayofreason]

48. Sorry..

...just because I did not use my knowledge of Shakespeare today doesn't mean that I shouldn't have had it in high school.

And while plumbers don't need to use calculus or newtonian mechanics in everyday life or their work doesn't mean that having a scientifically literate citizenry is a bad idea. I, for one, am strongly in favor of the Texas 4x4 plan. No one can with certainty predict the path that students will take, and that path should include the preparation needed to be successful in college if the student takes that route. Take a look at the recently released College Board standards as an example of what national experts think is required.

So, yes, not everyone needs to go to college (though statistically your earning potential is severely impacted if you don't), and not everybody needs to be a scientist or engineer (though statistically your earning potential is significantly higher if you are, and we need more scientists and engineers for our techo-economy to grow and provide the wealth society needs). However, anyone, in any profession, who is not scientifically literate is at the mercy of forces they cannot comprehend or control, and they will not be able to make informed decisions in an increasingly technological and scientific world.

[wtmusic]

21. Unless every kid's parents had a screaming match the night before

it wouldn't significantly affect a teacher's score.

I don't know how the LAT applied their "value-added" approach, but it was supposed to only measure improvement. So a difference in resources between districts wouldn't matter.

If electricity is cut off the week before a test, in CA kids have up to five days to make it up. If they're only teaching standards material a week before the test there's a problem anyway.

I don't agree with the Times' evaluation, putting a label of "not effective" on one simple metric. Too many parents will see the label and assume it is an overall evaluation. But it's a valid metric and should be used as part of a teacher's evaluation.

[mbperrin]

25. I had 3 kids out of 48 not pass the test. That gave me a 94% rate, the

absolute minimum. One kid was absent 102 days out of 145 in the testing period. Another was in the hospital having emergency surgery for a car wreck. The third simply failed. Perhaps it's worth mentioning that 25 of the 48 failed their 10th grade TAKS altogether, but that doesn't count in Texas. I'm a minimally effective teacher by their lights.

There are no makeups on the first test date, regardless. The student can make up their tests for graduation purposes, but the first score stands, and absent is a zero in Texas.

[wtmusic]

35. Is that based on the LATimes analysis, or Texas's? nt

[mbperrin]

39. That is the state of Texas system.

nt

[mbperrin]

43. If I had one more child fail, I would have a 92% pass rate, which is failing

in Texas. One more child for any reason, including simply missing the test.

So yeah, each child matters.

[wtmusic]

54. We're discussing two different subjects here.

One is the validity of using improvement in students' test scores to evaluate teachers. That's what the Times has done, arguably with a rush to judgement.

The other is the State of Texas' use of raw student test scores to evaluate teachers.

Though the Times has no right to post a rating of "Ineffective" for any teacher based on that one metric, it's valid as part of an evaluation.

The Texas system, if I understand it correctly, is off-the-charts stupid.

[mopinko]

49. one day, one test, one kid,

a blip. nobody is using one data point. more like one kid, one test, each year, times the total number of kids. gives you are pretty large data set, and results over time.
it shouldn't be all, but it should be a big part. half, or a third.

[mbperrin]

53. Sorry to bust your bubble, but this is the single data point.

Teachers were not rehired and others were moved over their results on this one test, one day.

[LWolf]

59. +1,000

[AguaAzul]

11. I am One of the 6,000 LA Teachers Scapegoated

I am one of those 6,000 teachers and while I was pissed off at first (and cancelled my subscription to the Times), I am now less stressed though in a quandary about how to proceed. Obviously, based on flawed data on an inaccurate measure of my effectiveness as an educator, I did not come out well in the database.

I have always worked in the inner city with English Language Learners. I have no way of knowing if wrong answers were due to the students' not understanding the question (due to limited English), lack of knowledge, or simply lack of interest. The Times wants to judge my ability on THAT?!?!

I have always thought educating children involved more than teaching to the test. I follow the COMPLETE curriculum and also strive to meet the socio-emotional and psychological needs of my diverse students. While most of my colleagues have had the assistance of paraprofessionals, I did not for ALL THE YEARS of the study (if you can call it that). This is not even getting into the teaching to the test and even cheating that goes on when nobody monitors the teacher during and after the test.

I could raise test scores if that is all they want (teach to the test only). My dad and wife suggest I start playing the game. I am left in a moral dilemna. Do I do what I believe is right or what will take the pressure off? Thankfully, I moved to a new school and a grade where I don't have to administer the state tests anymore. I welcome any feedback or questions.

[YvonneCa]

26. From another California teacher...

... :grouphug: I am now retired...the stress of this is too much. It make our job MORE difficult.

Good luck to you. I hope you will post this in GD. People need to know...

[BillH76]

44. You've got to start playing the game.

But you should also encourage your union leadership to fight this ridiculous crap, for all the reasons you mention in your post.

[mopinko]

60. even the value added analysis should not be taken as a total

measure of a teacher's worth. the point of the value added analysis is to statistically overcome the differences in the abilities of the students, and really compare one year's kids to the next. but step 2 is supposed to be that teachers, especially in the lower quartiles, are supposed to be FURTHER evaluated, to see what they need. maybe in your case, it means that your materials suck, or you need an aide, or maybe it even means that you are a great teacher who doesn't look good in the stats. the teachers in the top quartile are supposed to be further evaluated to see what makes them so good. maybe that teacher gives the kids cookies and milk every day, or uses color markers or who knows.

i have said before and i say again- no person can be or should be boiled down to a number. teachers, kids, admins, parents, nobody, nobody, nobody. but i do think this is a useful set of stats that should form the basis of teacher evaluation. i don't think this is something that needs to be published in the paper, but i sure think it should be used (lausd didn't use it, or even look at it) and given to parents.

[Sancho]

13. As so many have said before...

don't you think that your dentist should have a news report of how many cavities the patients in each office have...it's such an insane way to measure "effectiveness" that I'm amazed that newspapers or the public fall for such garbage. These complicated value-added formulas are GIGO, and most of them know there are hugh flaws with that sort of system.

If you want a good explanation of VA, watch the video at

http://www.youtube.com/watch?v=uONqxysWEk8

[boppers]

29. Do you visit your dentist every day for 6 hours?

Interesting video, highlighting the complexity of gathering good metrics, but the dentist analogy is deeply flawed.

[Sancho]

50. not really....

most teachers see 30 kids for 45 minutes a day...and the idea is that a dentist may be good at filling cavities, or work in a neighborhood with patients who have poor diets, etc. You can't judge the dentist's skill by the number of cavities. Same thing with all the multitude of confounding factors in test scores on once a year testing. There are lots of technical analogies with this problem and value-added models, and the dentist story is often used as a parallel. See
http://www.teachervision.fen.com/education-and-state/education-and-social-issues/4727.html
for one version...

[Llewlladdwr]

22. After perusing this thread I have a question.

Is there any valid methodology at all by which teacher performance can be measured and rated?

[YvonneCa]

27. I'm sure there is, but it has yet to be...

...implemented. Until it is, this is unfair. The current system is flawed.

[boppers]

31. Depends on the subject, and what you're trying to measure.

Take, oh, an art teacher. It's easy to quantify if, after a year with the teacher, a student can identify common art materials, or identify historical art persons and styles. It's *hard* to quantify if the student has become a better artist themselves.

Extending that thought to reading, it's easy to quantify whether or not a student can paraphrase a paragraph. It's much harder to test whether or not a student actually understands the deeper meanings and subtexts within a paragraph. In math, it's easy to test if a student has memorized, and can utilize, formulas. It's harder to test if they actually innately understand the formulas, or can invent new ones on their own. In just about any area of study, once you get past basic memorization and simple transformation, testing becomes very difficult.

Where I am in favor of testing, is for testing base standards that *can* be quantified, the bare minimums to be met, so that an art student could identify the difference between paint and charcoal.... but testing is a horrible measure of who's going to be a good artist, and for that matter, who is a great teacher beyond meeting those bare minimums.

[primavera]

56. Doubtful

Professional prowess depends upon so many subtle and ineffable qualities, standardized tests that attempt to reduce everything down to a number are seldom up for the challenge. Tests may be able to reflect a student's capacity for rote memorization, but rote memorization doesn't make an Einstein or a Mozart or a van Gogh; it makes a pocket calculator.

[wtmusic]

63. If "valid" means "flawless", then no.

However...

Attendance, punctuality, student satisfaction, parent satisfaction, student test score improvement, peer evaluation, and administrator evaluation can be combined to produce a relatively objective measure of whether a teacher is doing his/her job, or not.

[BillH76]

32. Flaw in Value Added Testing: schools put kids in

"tracks."

In a paper published in February in the Quarterly Journal of Economics, Princeton University economist Jesse Rothstein uses some sophisticated modeling techniques to suggest that such techniques could be based on shaky assumptions.

Using student-testing data from North Carolina, Rothstein makes his case by developing a "falsification" test for value-added models. For example, he wondered, would the model show that 5th grade teachers have effects on their students' test scores in 3rd and 4th grades? Since it's impossible for students' future teachers to cause their previous achievement outcomes, Rothstein reasons, there should be no such effects.

But in fact there were—and they were quite large. Rothstein says this happens because students are not randomly sorted into classrooms. A principal, for example, might assign a large number of students with behavior problems to a teacher who is known to have a way with problem students or parents of high achievers might lobby to get their child in a class with the "best" teacher. When that happens, though, it biases the results of value-added calculations.

I went to Allesandro Elementary School, which gets very low marks on overall value-added. By contrast, Ivanhoe School, a couple of miles away, gets great marks. The difference? The neighborhood. Allesandro is in an area of latch-key poverty stricken non-English speaking kids. The school was one big behavioral problem when I went there, and I'm sure it still is. Ivanhoe, on the other hand is in a much more affluent neighborhood, and there were hardly any behavioral problems when my brother went there, and I'm sure it's still the same. Of course the rate of improvement will be affected by this, as it will if you put the latch-key misbehavers in one class and the upper middle class kids in another. The latch-key teacher will not do as well as the teacher of the upper middle class kids.

[boppers]

33. A more detailed study:

http://isites.harvard.edu/fs/docs/icb.topic245006.files/Kane_Staiger_3-17-08.pdf

In short, you can normalize for "tracks", and a whole host of other problems, with much more sophisticated number crunching. That being said, I'm not sure how deeply the LAT numbers were analyzed.

[BillH76]

37. The kids in the study you cite were not "tracked"

From the article: "In this paper, we rely on an experiment in which 78 pairs of classrooms (156 classrooms and 3194 students) were randomly assigned between teachers in the school years 2003-04 and 2004-05."

http://isites.harvard.edu/fs/docs/icb.topic245006.files/Kane_Staiger_3-17-08.pdf

[boppers]

51. It's a little more complex than that.

First, to establish a comparison baseline, they started with a baseline that was not random: The set of teachers, and their prior history. Then they compared those past, possibly tracked, performance predictions to outcomes with a random assignment.

So, in essence, they were comparing tracking with randomness (among many other things).

[Hissyspit]

57. Where are the administrator effectiveness ratings?

Where are the standardized school board effectiveness ratings?

[TomCADem]

58. What's the difference between this and wikileaks? Transparency is good!

Public teachers are public employees like members of the military, police officers, etc. Why should we allow the government to hide the performance of teachers? More disclosure and transparency!

[wtmusic]

61. Yes and no.

Although the numbers themselves should be public, the Times assigned a randking ("Not effective", etc) based on a single metric.

Misleading and counterproductive.