News for hackers. Stuff that matters.
Re:The Case for $320,000 Teachers by yummyfajitas on Wednesday July 28, 2010 @11:10AM (#1554499)
If you object to standardized tests, what do you propose we use? Non-standard tests? The opinions of assorted cronies?
Re:The Case for $320,000 Teachers by jbooth on Wednesday July 28, 2010 @11:16AM (#1554520)
Well, that's sort of like saying we should use lines of code as a metric, because what else to we have to go on. The opinions of assorted developers?

There needs to be some attempt at normalization between teachers and school districts, of course. But there's room for qualitative measures instead of quantitative ones. Especially when the tests seem (unintentionally) designed to punish teachers for teaching poor/minority/ESL students.

Re:The Case for $320,000 Teachers by yummyfajitas on Wednesday July 28, 2010 @11:34AM (#1554571)
We have much better metrics for developers than lines of code - bug fixes/feature requests weighted by importance, P&L, etc. Do you have a better metric for teachers than standardized tests?

Especially when the tests seem (unintentionally) designed to punish teachers for teaching poor/minority/ESL students.

This is utterly false if we reward teachers based on deltas rather than absolute values, that is we compare actual performance to statistically predicted performance and reward/penalize teachers when actual performance exceeds predicted performance.

Re:The Case for $320,000 Teachers by kingkilr on Wednesday July 28, 2010 @12:14PM (#1554709)
> bug fixes/feature requests weighted by importance

Isn't there a DailyWTF about a company that did just that and found that the testing people conspired with the developers to invent bugs, locate them, and then collect bonuses for fixing them.

Re:The Case for $320,000 Teachers by yummyfajitas on Wednesday July 28, 2010 @01:25PM (#1554947)
Preventing conspiracies like this is exactly why teachers should be evaluated by standardized tests designed by an independent agency.
Re:The Case for $320,000 Teachers by jbooth on Wednesday July 28, 2010 @01:46PM (#1555028)
You already said you think teachers are overpaid and underworked and your goal is to make things harder for them.

Why should they want to work under a system like that?

What independent agency would you want devising metrics to figure out how good you are at your job while knowing nothing about the specific challenges you deal with?

Re:The Case for $320,000 Teachers by yummyfajitas on Wednesday July 28, 2010 @03:10PM (#1555297)
I believe teachers currently are comped at above-market rates. I would bring them down to market rates. They would work under that system for the same reason most other professionals work at market rates - it's the best they can get.

The independent agency which currently devises metrics to figure out how good I do my job is the stock market. My bonus = % of profits.

When I was a professor, I favored (and suggested to the department) that some higher authority should devise midterm and final exams. It would a) allow us to measure teacher performance, as opposed to student opinion [1] and b) allow comparison of student performance across classes.

[1] To get good student evaluations, all you need is easy tests + jokes + good looks.

Re:The Case for $320,000 Teachers by jbooth on Wednesday July 28, 2010 @03:25PM (#1555341)
You've got market and non-market all mixed up.

Public education isn't a market good, it's a public good. Toilet paper and oatmeal are market goods.

The labor market for teaching is in fact a free market. Any given municipality can pay whatever they want to pay, and get the talent they're paying for accordingly. The fact that teachers are getting laid off in something like 80% of municipalities the last few years means you don't even have to scrape the bottom of the barrel.

Agree on the statements about professors. All the same. If you were working on something more qualitative as part of a team, and not a lone trader, we both know there's no quantitative way to assess your work. Lines of code is a joke and other metrics invariably bring in subjective measures.

EDIT to add: If you're pro-free market, shouldn't it be up to municipalities to choose which teachers they want to keep or fire as they see fit, without some bureaucrat telling them that according to some BS test scores they have to do such and such?

Re:The Case for $320,000 Teachers by yummyfajitas on Wednesday July 28, 2010 @03:45PM (#1555405)
Education is not a public good since it's both rivalrous and excludible. That makes it a private good. The fact that a private good is provisioned by the government does not make it a public good.

Also, the labor market for most govt employees is not remotely a free market. Unions keep out the competition, essentially getting no-bid contracts to supply labor. They play politics to elect politicians who will give them what they want - in many cities the teachers unions essentially control the political system. You might as well call Halliburton's no-bid contracts in Iraq a "free market". It's a "free market", closed to new entrants, with no negotiation between the buy and sell side, but with the buy and sell side colluding to rip off the public.

In any case, just because it's tough to quantitatively measure some development does not mean it's tough to measure teacher performance. Students either know the material at the end of the term or they don't, simple as that.

All the whining about how it's impossible is really just an attempt by teachers unions to avoid accountability. I'm sure Halliburton would love to avoid accountability for their performance in Iraq and elsewhere too, doesn't mean we should take them seriously.

Re:The Case for $320,000 Teachers by jbooth on Wednesday July 28, 2010 @04:09PM (#1555494)
So, by your logic, roads and sewage are private goods? Basically everything except the military? Consistent with a libertarian.

In almost no city does the teacher's union essentially control the political system. The police union is always, always more powerful, for one. They're just more politically engaged, people who are driven by power. Several private unions will be more powerful. And that's leaving out every power broker that isn't a union (representing most of the power).

"All that whining about how impossible it is to measure programmer productivity is just an attempt by developers to avoid accountability." You might consider that you don't know a single thing about teaching in a non-college or lower-class environment. No offense. I don't claim to, either. But I'll at least listen to what every teacher I've ever met says about the matter.

Re:The Case for $320,000 Teachers by shasta on Wednesday July 28, 2010 @07:59PM (#1556072)
Programmers may routinely be judged by peers, but the entire team is then judged by results. A team that can't ship gets the axe or goes under. The closest counter-examples to this I can think of, doing government contracts, have also been some of the worst software shops that I've encountered.

So what's the analogous mechanism going to be? Are you ok closing entire schools if they have lower than average test scores in their class (whatever this means)? Otherwise, how do prevent the whole system from devolving into politics and cronyism?

Re:The Case for $320,000 Teachers by kingkilr on Wednesday July 28, 2010 @01:55PM (#1555059)
This presupposes that a standardized test can test anything of value. I've spent the majority of my last 16 years in school, since the 3rd grade or so I've had no problems evaluating the competence of my teachers, and I didn't need a test to do so.
Re:The Case for $320,000 Teachers by dhimes on Wednesday July 28, 2010 @02:30PM (#1555184)
But how do you know your evaluation is accurate? A colleague of mine was fond of saying that asking your average student what she thought of her teacher was like asking your average seven year old what she thought of her flu shot.

Even if you are exceptional in this regard, the average person will not do well. And I have read a lot of reviews in my life from college students.

Until you know the subjects well, you are only evaluating whether or not you liked it.

Re:The Case for $320,000 Teachers by abstractbill on Wednesday July 28, 2010 @11:49AM (#1554627)
This is utterly false if we reward teachers based on deltas rather than absolute values...

That's a big if. From what I've seen, this is practically never done - absolute values are the norm.

Re:The Case for $320,000 Teachers by lutorm on Thursday July 29, 2010 @04:09PM (#1559065)
I don't have an alternative that I think is workable. And that's why I'm far from convinced that teacher salaries should be tied to performance; I think the side effects of "optimization by proxy" will be severe.

As an aside, pointing out that something is bad does not put the onus on me to invent something better. The sensible conclusion is to not do anything until something good has been worked out. The alternative is to do something just for the sake of appearing to do something.

Re:The Case for $320,000 Teachers by ewanmcteagle on Wednesday July 28, 2010 @11:38AM (#1554589)
If students weren't locked into schools enrollment levels and parent feedback would provide a lot of information.
Re:The Case for $320,000 Teachers by ktsmith on Wednesday July 28, 2010 @02:22PM (#1555160)
Parent feedback isn't worth nearly as much as you might think. Parents aren't in the classroom, they tend to get one side of things (their child's). Some parent's are objective and some aren't. This leads to plenty of horror stories about parents that assume all teachers are morons and their children are infallible, I've seen it first hand and it's really pathetic. The flip side of that would be the parents of my wife's former students. She has been teaching in poor neighborhoods and a significant percentage of the parents are illiterate. While these parent's are very happy with my wife and they request all their children (and cousins etc) have her, using their feedback to rate my wife as a teacher would be seriously flawed as they really don't have the ability to judge if the teaching is effective at all.