Massive Cheating Scandal Rocks Atlanta Schools

[Traitorfish]

Meaning: Your support of education shouldn't depend on your ideology. It depends on whether you want to see humanity advance or not.

But that is by definition an ideological decision; indeed, even the concept of "human advancement" must be determined within an ideological framework. Whiggism has a fairly hegemonic grip these days, but its still an ideology, as much as Marxism, fascism, or any other.

I'd also like to ask you to explain TF's comments "slapped me in the face".

As he said, it wasn't intended seriously, but I'm assuming that what he meant was my reference to, in effect, the proletarianisation of teachers. To save getting into a long-winded rant, what that basically means is the demotion of teachers from a somewhat ambiguous middle strata, enjoying certain functional autonomy, to that of the standard prole (a white collar never having been a true barrier to entry); the thorough subjugation of this particular field of labour to capital (in this case, state capital).

 

[civ_king]

Can't they use short answer questions?

No, because machines don't do short answer questions

 

[SS-18 ICBM]

But why does it have to be a machine that does the marking? Don't elementary school teachers have enough time? Besides, I think such answers better reflect learning than shading circles.

 

[civ_king]

But why does it have to be a machine that does the marking? Don't elementary school teachers have enough time? Besides, I think such answers better reflect learning than shading circles.

Middle and High School teachers don't have the time

 

[downtown]

I give any teacher the default understanding that they putting their best effort out to educate their students and are capable of doing so, unless proven otherwise. In this particular case, teachers cheating on exams =/= teachers not adequately educating their students, so I would be very careful about mass firings at the classroom level. However, it is a complete failure on the part of educational administrators and I would be going at them with a pretty big broom.

Yeah, given the pressure put on the teachers by the administrators, I think most of the blame goes with them.

Yep. Though there are other ways to sort of game the system. Mis-report drop outs, for example. Or have the state run it's own benchmark program. IIRC both were implicated in the "Texas Miracle", which pretty directly led to the NCLB act... which is likely much of the reason for the Atlanta scandal.

Right, but I think those are slightly different things (one is legal, one isn't). It's true though, that getting really solid and accurate student achievement data is really hard.

But why does it have to be a machine that does the marking? Don't elementary school teachers have enough time? Besides, I think such answers better reflect learning than shading circles.

You need standardized tests to be graded as standardized as possible. Short answer questions are WAY more expensive and take more time to grade, and also have higher error rates. You obviously wouldn't want a district employee to be the one who grades the tests (the temptation to cheat would be way too large), so they have to be sent to state officials, or outsourced completely. The tests for my students were graded by somebody none of us had met, 100 miles away.

 

[Yui108]

This is basically the same thing as grade inflation at top highschools to abuse scholarships.

I'm curious about this one. Maybe you're referring to prep schools or something else, but I attend the best, or second-best (dependent on ranking methodology) public high-school in Illinois, and grade deflation is much more of an issue than grade inflation. It is much harder to earn a 4.0 at my school than in (most of) the CPS or even a less affluent suburban school.

 

[warpus]

It seems that some school districts are totally uninteresting in actually educating their students.

Of course not! They aren't instructed to teach, they're instructed to make sure that their students get good marks.

There's a huge disconnect there.

So I gotta ask.. What's going to happen to all those people (teachers, principals, etc.) who participated in this? They are all going to be fired, right?

 

[Neonanocyborgasm]

Seems like a likely outcome of the perverse incentives of NCLB. What should be done is give enough funding to ALL school districts to run schools adequately, and THEN remove those staff that are significant underperformers.

 

[Japanrocks12]

This sucks, but the only way to completely eliminate stuff like it is to revamp the entire educational system, from middle school all the way up through higher education programs (grad, med, law, whatever) schools to make grades and GPA count for far less than it currently does. Which is a thought that the American educational system, which cares more for bureaucracy and efficiency in lieu of actually providing a complete and proper education, will never entertain.

 

[downtown]

I'm curious about this one. Maybe you're referring to prep schools or something else, but I attend the best, or second-best (dependent on ranking methodology) public high-school in Illinois, and grade deflation is much more of an issue than grade inflation. It is much harder to earn a 4.0 at my school than in (most of) the CPS or even a less affluent suburban school.

I think it's typically more of a problem at elite universities and private high schools than elite public ones. I'm not sure if there has been an academic study on it at the public K12 level or not.

Seems like a likely outcome of the perverse incentives of NCLB. What should be done is give enough funding to ALL school districts to run schools adequately, and THEN remove those staff that are significant underperformers.

Yeah, the HUGE amount of money (and prestige) attached to test results from NCLB and RttT are certainly part of the reason for stuff like this. It's quite possible we'll have another bombshell like this over the summer, since the Yahoo News story said 5 other states had districts with similar testing abnormalities.

This sucks, but the only way to completely eliminate stuff like it is to revamp the entire educational system, from middle school all the way up through higher education programs (grad, med, law, whatever) schools to make grades and GPA count for far less than it currently does. Which is a thought that the American educational system, which cares more for bureaucracy and efficiency in lieu of actually providing a complete and proper education, will never entertain.

Well, this sort of this doesn't have anything really to do with grades, just standardized test scores. Stopping teacher/admin led GRADE inflation would be nearly impossible to do on a grand scale. You're going to have somebody audit every homework assignment? That increases teacher paperwork by an order of magnitude, and will make good instruction even harder.

 

[Japanrocks12]

Standardized test score and grade inflation via cheating or less illicit means are just two different sides of the same coin, the phenomenon of good stats at any cost being prioritized over the providing of a well-rounded education.

 

[downtown]

The Department of Education is launching a similar investigation in Washington DC. 2 teachers have already been fired in DC for cheating, and the scope of cheating there is highly unlikely to be the scope of Atlanta, it could still be more significant.

http://www.washingtonpost.com/local.../07/gIQAK2Cg2H_story.html?wprss=rss_education

 

[Neonanocyborgasm]

Yeah, the HUGE amount of money (and prestige) attached to test results from NCLB and RttT are certainly part of the reason for stuff like this. It's quite possible we'll have another bombshell like this over the summer, since the Yahoo News story said 5 other states had districts with similar testing abnormalities.

I wouldn't be surprised. NCLB seemed to have "perverse incentives" written all over it from the start. It also makes a rather insulting presumption about teachers, which is that they mostly are damned incompetent and lazy, and are the cause of educational deficiency in the country, and that if only we rewarded the good and punished the bad (by witholding reward), we could sort them out. Another faulty presumption of NCLB is that money fixes education, but if that's true, why be selective? Why not just give everyone money and education will be fixed everywhere? The logical inconsistency is so mind-boggling, it could only have been fabricated in Congress.

 

[pi-r8]

Using test scores as the only measure of how good a school is, is a terrible idea. Combine that with merit pay for high test scores and you're giving people a huge incentive for cheating, like what happened here. Maybe the administrators decided that cheating was better than having the school lose all it's funding because of bad test scores.

Unfortunately, I don't know any way to objectively measure school performance without looking at test scores. Hiring more teachers, and giving them a little more trust and autonomy, while making sure that students have many different teachers so they don't get stuck with just one bad teacher, would probably be the best method.

 

[Bestbank Tiger]

Seems like a likely outcome of the perverse incentives of NCLB. What should be done is give enough funding to ALL school districts to run schools adequately, and THEN remove those staff that are significant underperformers.

That raises two other questions:

1) Will actual underperformers cheat on the test, and if so will that hurt honest teachers who end up getting fired because they look like underperformers?

2) How do you define an underperformer? Low standardized test scores could be a reflection of the teacher from the current year, the teachers over the past 3-4 years, or the parents and the home environment.

 

[downtown]

Unfortunately, I don't know any way to objectively measure school performance without looking at test scores. Hiring more teachers, and giving them a little more trust and autonomy, while making sure that students have many different teachers so they don't get stuck with just one bad teacher, would probably be the best method.

It isn't rocket science. Using value added test scores is a perfectly valid metric for PART of the evaluation process, but the biggest should prob be classroom observations, by either administrators or master teachers. Teachers should also be judged on their ability to meet other professional development goals...like most professionals. We don't judge doctors JUST on their mortality rate, or journalists JUST on pageviews.

That raises two other questions:

1) Will actual underperformers cheat on the test, and if so will that hurt honest teachers who end up getting fired because they look like underperformers?

I think the ATL scores are all getting thrown out, but if a district had massive cheating, I could see that being an issue.

2) How do you define an underperformer? Low standardized test scores could be a reflection of the teacher from the current year, the teachers over the past 3-4 years, or the parents and the home environment.

Typically, you use something called value added data, where you measure the student's growth that year. If a 5th grader enters your classroom at the 2nd grade level, and leaves at the 4th grade level, he's going to fail his end of year assessment, but you're still a very good teacher (growth above 1 grade level a year is very good). Judging a teacher just based on pass/fail rates is pretty stupid.

Ideally, teachers are given a multiple year window (at least 2) before they're fired or otherwise penalized, because 1 year could be full of outliers (like you said).

 

[Neonanocyborgasm]

That raises two other questions:

1) Will actual underperformers cheat on the test, and if so will that hurt honest teachers who end up getting fired because they look like underperformers?

Who said there'll be a test?

2) How do you define an underperformer? Low standardized test scores could be a reflection of the teacher from the current year, the teachers over the past 3-4 years, or the parents and the home environment.

A teacher's performance is reflective of a persistent trend in the students, not a single test. A single test can be skewed by a some bad students in the class, but consistently poor performance, especially poorer than others, cannot be an accident. I would say that if a teacher's class performs consistently worse than some standard, perhaps over 4 cycles, then that is cause to doubt the teacher's ability. Even so, it is only doubt, not conviction.

As you can see, I don't favor witch-hunting teachers for the poor performance of the system as a whole. I favor fixing the system, not scapegoating some politically expedient pee-ons.

 

[Mr. Dictator]

I'm more interested in how this culture of cheating was able to fester. With the cheating parties, it was obviously a very open secret. I have a hard time believing everybody was completely fine with it though.

 

[downtown]

I'm more interested in how this culture of cheating was able to fester. With the cheating parties, it was obviously a very open secret. I have a hard time believing everybody was completely fine with it though.

I don't think it's a big mystery. Administrators threatened whistleblowers with their jobs. There are more teachers than job openings nationwide, and the South (and Georgia in particular) have practically no teachers unions. The message was clear...cheat or get fired. Not a big surprise that a lot of folks cheated.

 

[downtown]

Aaaaand one more time! Now the cheating inquiry has been extended to dozens of schools in Pennsylvania, mostly around the Philly region.

How many more are going to pop up?

http://www.thenotebook.org/blog/113871/2009-report-identified-pa-schools-possible-cheating