BTW you see a similar effect, though for different reasons, in higher education - college costs are going up and they're largely being driven by administrative, non-teaching-staff, and facilities expenditures.
While that's part of it, I would contend that a larger part of what's driving up college costs are the continual cuts in state funding.
Yup. It's a pretty easy correlation to prove - I've seen the graphs. Salaries - even the much touted administrator salaries - only account for a fraction of the higher tuition. The lion share comes from state legislators having substantially cut funding to public colleges. In some states, the share made up by state appropriations is less than a quarter of what it was in the 1980s.
Not really the same thing.
When I say 'college costs' I'm talking about total operational costs. I'm not talking about tuition prices.
You guys are talking about who foots the bill. Does the state foot the bill through subsidies, or do students foot the bill through loans?
That's a good discussion to have, but the deeper underlying question is - why has the bill gotten so big? You can't answer that question by looking at funding sources.
Feral on
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
I ran into an interesting article on school spending vs school quality broken down to a district level. It talks about the lack of correlation between spending and performance, but it also brings up mention of specific spending issues that lend some support for what I was wondering earlier.
In particular, he found that urban schools tended to spend inefficiently for a variety of reasons, including high staff and student turnover and conflicts over how to teach struggling students. At the same time, he said, urban districts often have extra expenses for needs such as security, dropout prevention, or for teaching students who are not proficient in English.
Deanne Fitzmaurice/California WatchTeacher Mellissa Schmitz high-fives a student at Costano Elementary School in Ravenswood City School District, East Palo Alto.
One of those districts with higher expenses is the Ravenswood City School District in East Palo Alto, where about two-thirds of students are English learners. The district has had to hire three full-time Spanish translators – mainly to translate lengthy special education reports as required by law – and has translators working in the school office, in classrooms and at parent meetings.
Ravenswood spends nearly $13,000 per student, yet has cut several programs and may slash two weeks from the next school year, said Superintendent Maria De La Vega.
The correlation isn't perfect, and there are a few counter-instances involving frugal and efficient use of dollars, but what I'm getting from the article is that a lot of the issue is that schools are not getting the funding they might need on a per-school basis to address their specific needs. And that, specifically, you can't just look at the final per-student spending for a school as if that is the final word on whether that school is adequately funded.
I submitted an entry to Lego Ideas, and if 10,000 people support me, it'll be turned into an actual Lego set!If you'd like to see and support my submission, follow this link.
BTW you see a similar effect, though for different reasons, in higher education - college costs are going up and they're largely being driven by administrative, non-teaching-staff, and facilities expenditures.
While that's part of it, I would contend that a larger part of what's driving up college costs are the continual cuts in state funding.
Yup. It's a pretty easy correlation to prove - I've seen the graphs. Salaries - even the much touted administrator salaries - only account for a fraction of the higher tuition. The lion share comes from state legislators having substantially cut funding to public colleges. In some states, the share made up by state appropriations is less than a quarter of what it was in the 1980s.
Not really the same thing.
When I say 'college costs' I'm talking about total operational costs. I'm not talking about tuition prices.
You guys are talking about who foots the bill. Does the state foot the bill through subsidies, or do students foot the bill through loans?
That's a good discussion to have, but the deeper underlying question is - why has the bill gotten so big? You can't answer that question by looking at funding sources.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
BTW you see a similar effect, though for different reasons, in higher education - college costs are going up and they're largely being driven by administrative, non-teaching-staff, and facilities expenditures.
While that's part of it, I would contend that a larger part of what's driving up college costs are the continual cuts in state funding.
Yup. It's a pretty easy correlation to prove - I've seen the graphs. Salaries - even the much touted administrator salaries - only account for a fraction of the higher tuition. The lion share comes from state legislators having substantially cut funding to public colleges. In some states, the share made up by state appropriations is less than a quarter of what it was in the 1980s.
Not really the same thing.
When I say 'college costs' I'm talking about total operational costs. I'm not talking about tuition prices.
You guys are talking about who foots the bill. Does the state foot the bill through subsidies, or do students foot the bill through loans?
That's a good discussion to have, but the deeper underlying question is - why has the bill gotten so big? You can't answer that question by looking at funding sources.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
Scholarships have also been going up, which is likely a good thing.
BTW you see a similar effect, though for different reasons, in higher education - college costs are going up and they're largely being driven by administrative, non-teaching-staff, and facilities expenditures.
While that's part of it, I would contend that a larger part of what's driving up college costs are the continual cuts in state funding.
Yup. It's a pretty easy correlation to prove - I've seen the graphs. Salaries - even the much touted administrator salaries - only account for a fraction of the higher tuition. The lion share comes from state legislators having substantially cut funding to public colleges. In some states, the share made up by state appropriations is less than a quarter of what it was in the 1980s.
Not really the same thing.
When I say 'college costs' I'm talking about total operational costs. I'm not talking about tuition prices.
You guys are talking about who foots the bill. Does the state foot the bill through subsidies, or do students foot the bill through loans?
That's a good discussion to have, but the deeper underlying question is - why has the bill gotten so big? You can't answer that question by looking at funding sources.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
Scholarships have also been going up, which is likely a good thing.
Yes and no.
It's good that scholarships are going up, but this is another reflection of the rising burden on the students and families. A lot of scholarships are funded by tuition, so the burden is just spread out a bit.
And that's also coming under attack. There's a strong move on in conservative circles to curtail or outright ban using tuition monies to pay for scholarships. They say its not fair to force families to subsidize poor students. This is definitely a bubbling talking point, originating from think tanks like the Civitas Institute.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
I don't know why you're linking a chart on US government spending when that is largely irrelevant to what I'm talking about.
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
I don't know why you're linking a chart on US government spending when that is largely irrelevant to what I'm talking about.
Quick Googling. I need to find better data, but I've been to enough presentations on the subject to know that the basic idea is the correct - college costs aren't skyrocketing anywhere near to the point that the political rhetoric makes out. It's just that students and parents are shouldering more of the burden.
There's fat to be cut in administration - although a lot of that administrative growth has arisen from the need to develop and maintain additional funding sources. There's fat to be cut in dorms and on-campus dining - although cutting back runs into free market problems in that schools have to maintain housing and food choices but anything below a certain level of quality will not be competitive with the private sector, students will eat/live elsewhere and the lack of profits will end up costing schools more money to maintain.
The issue is actually pretty simple. States have cut, on average, 75 percent of their funding of public schools. The federal government is fickle as well, cutting back grants in favor of loans. Colleges are expensive to maintain and that money has to come from somewhere.
But the reality of this conversation has been drowned out in the bullshit.
BTW you see a similar effect, though for different reasons, in higher education - college costs are going up and they're largely being driven by administrative, non-teaching-staff, and facilities expenditures.
While that's part of it, I would contend that a larger part of what's driving up college costs are the continual cuts in state funding.
Yup. It's a pretty easy correlation to prove - I've seen the graphs. Salaries - even the much touted administrator salaries - only account for a fraction of the higher tuition. The lion share comes from state legislators having substantially cut funding to public colleges. In some states, the share made up by state appropriations is less than a quarter of what it was in the 1980s.
Not really the same thing.
When I say 'college costs' I'm talking about total operational costs. I'm not talking about tuition prices.
You guys are talking about who foots the bill. Does the state foot the bill through subsidies, or do students foot the bill through loans?
That's a good discussion to have, but the deeper underlying question is - why has the bill gotten so big? You can't answer that question by looking at funding sources.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
Scholarships have also been going up, which is likely a good thing.
Yes and no.
It's good that scholarships are going up, but this is another reflection of the rising burden on the students and families. A lot of scholarships are funded by tuition, so the burden is just spread out a bit.
And that's also coming under attack. There's a strong move on in conservative circles to curtail or outright ban using tuition monies to pay for scholarships. They say its not fair to force families to subsidize poor students. This is definitely a bubbling talking point, originating from think tanks like the Civitas Institute.
You can get around that by shuffling like 4 lines on a financial statement. Seems like a stupid thing to go after.
BTW you see a similar effect, though for different reasons, in higher education - college costs are going up and they're largely being driven by administrative, non-teaching-staff, and facilities expenditures.
While that's part of it, I would contend that a larger part of what's driving up college costs are the continual cuts in state funding.
Yup. It's a pretty easy correlation to prove - I've seen the graphs. Salaries - even the much touted administrator salaries - only account for a fraction of the higher tuition. The lion share comes from state legislators having substantially cut funding to public colleges. In some states, the share made up by state appropriations is less than a quarter of what it was in the 1980s.
Not really the same thing.
When I say 'college costs' I'm talking about total operational costs. I'm not talking about tuition prices.
You guys are talking about who foots the bill. Does the state foot the bill through subsidies, or do students foot the bill through loans?
That's a good discussion to have, but the deeper underlying question is - why has the bill gotten so big? You can't answer that question by looking at funding sources.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
Scholarships have also been going up, which is likely a good thing.
Yes and no.
It's good that scholarships are going up, but this is another reflection of the rising burden on the students and families. A lot of scholarships are funded by tuition, so the burden is just spread out a bit.
And that's also coming under attack. There's a strong move on in conservative circles to curtail or outright ban using tuition monies to pay for scholarships. They say its not fair to force families to subsidize poor students. This is definitely a bubbling talking point, originating from think tanks like the Civitas Institute.
You can get around that by shuffling like 4 lines on a financial statement. Seems like a stupid thing to go after.
In most public schools, a certain amount of scholarships are cooked into everyone's base tuition. Every time tuition has gone up, the universities here have made a point of noting that some percent (usually 15-25 percent) of the increase will go toward need-based scholarships. That's gotten the conservatives in a dander, since its government redistributing their wealth.
Let's say you want to improve our nation's roads. So you define a single metric to measure road quality: Number of potholes per 100 yards of road. That is the metric. Rural roads, urban roads, interstate highways, two lane roads, ten lane roads, they all get the same metric. And if your road has too many potholes? Well, you clearly aren't doing your job well, Mr. Road Maintainer, so we're going to cut your funding until you make those potholes go away. Clearly that will give you the incentive to do your job right.
Oh, okay, you're right, that is too rigid a metric. Let's add some variations to our criteria that account for road-width and maybe proximity to high-population centers. But still, it's just going to have to do with potholes, because that's an easy thing to measure, whereas all the other things that define how good a road is are just too hard to gauge. What, you don't like that metric? I guess you're aligned with those lazy road maintenance unions!
And that's basically the conversation we're having, as a nation, over grading our schools. Test scores are an important metric, but they are not the only important metric, and judging teachers based strictly on those kinds of numbers is asinine. Judging a school based strictly on those numbers is asinine. Those numbers are useful and important in determining where we need to focus our efforts, but they are not an indicator of quality in the way that diehard testing proponents want them to be.
Which isn't to say that we shouldn't grade teachers on an individual level. But we should trust the individual schools to hire and fire teachers based on their own criteria, just as we trust any other publicly run professional industry to do the same. Cutting funding to underperforming schools based on low scores is like cutting funding to the law enforcement budget because crime is going up. It's stupid and reactionary, and that's why people are opposing the implementation of these metrics - not because they don't care about accountability, but because the conversation is about very narrow metrics applied in stupidly counterproductive ways.
OR apply it to cops. Should we grade cops by how many arrests they make? That's their job right?
Except some students have professor and scientist parents areas are already fairly crime free. So do you grade this cop down because he's not making improvements arresting people? Or do you grade him higher than the cop in the inner city because improving from the bottom is easiermore crimes go unsolved there? Is a result or change of results arrest or change in crime the same everywhere?
You end up spending so much time money and energy not only trying to come up with a reasonable metric and evaluation regime but figuring out what to do with those measurements that you end up in a huge hole before you even start making improvements.
I'm wondering if the reason Finland's educational outcomes are superior to the US's despite per-capita dollars spent has more to do with their vastly reduced population density and economic disparity than any inherent superiority in their teaching methods or funding.
Maybe the dollars it takes to educate kids goes up exponentially rather than linearly as population density and economic disparity increases. Helsinki, I believe, is Finland's most densely populated city with a whopping 2700 people per sq. km. Then you look at New York City, which has about 10,600 people per sq. km. Not to mention the fact that in New York, 1% of the population controls 1/3 of the wealth (according to adjusted federal taxable income). I don't know what Helsinki's economic disparity is, but Finland's national economic disparity is about a third of what it is for the US.
Maybe we shouldn't be asking "how do we fix education" but rather "how do we curb the economic factors that hurt education".
Basically I'm just reiterating what was brought up on the first page:
The education problem in this country is actually a poverty problem. Full stop.
The issue isn't even school funding per se, although that is a confounding factor in many districts. The problem is that children in poor neighbourhoods do not come to school prepared to learn. The achievement gap between rich and poor and black and white is evident before formal schooling even begins. This is the result of a constellation of problems that plague poor children. These problems include but are not limited to lack of intellectual stimulation in the home during critical periods of brain development, malnutrition, health problems resulting from lack of access to care, health problems resulting from exposure to environmental toxins, and parents who are themselves poorly educated and do not know how to support their children's learning at home.
Research does show that teacher quality is the most important factor in a child's education within the school itself. However, socioeconomic status is overwhelmingly more important to a child's success in school than anything that happens in the classroom.
Quick Googling. I need to find better data, but I've been to enough presentations on the subject to know that the basic idea is the correct - college costs aren't skyrocketing anywhere near to the point that the political rhetoric makes out.
It's just that students and parents are shouldering more of the burden.
There's fat to be cut in administration - although a lot of that administrative growth has arisen from the need to develop and maintain additional funding sources.
Total education costs per student per year have been increasing faster than inflation or CPI since 1970. This is a noncontroversial statement. It's also not a new trend - it's a 40-year trend that has only recently come into the public eye due to the withdrawal of public funding sources. The public and the media didn't care nearly as much about the rising cost of education as long as the states and the feds were footing the bill.
So I guess if you're arguing against the idea that there's been a sudden 10-year spike in education costs... well, I'd agree with you there. That's been mostly driven by the withdrawal in state funding.
But I think the evidence is clear that there's also a longer-term, slower trend.
As for the need to develop and maintain additional funding sources, I agree that this is a problem, but I am skeptical that this is a majority cost driver. The book Tuition Rising breaks down in excruciating detail the expenditures of multiple institutions, and it shows that the need to hire more people to manage increasingly complex regulations (to secure funding, among other things) is a driver of personnel costs but it's not a majority one.
Here's the problem - simply restoring public funding to pre-1999 (or whatever recent year you choose) levels just papers over the problem. There are perverse economic incentives at play that stimulate spending on non-educational expenses and deincentivize cost cutting. Increasing student aid just pours more money into a cracked vessel.
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
I'm wondering if the reason Finland's educational outcomes are superior to the US's despite per-capita dollars spent has more to do with their vastly reduced population density and economic disparity than any inherent superiority in their teaching methods or funding.
Maybe the dollars it takes to educate kids goes up exponentially rather than linearly as population density and economic disparity increases. Helsinki, I believe, is Finland's most densely populated city with a whopping 2700 people per sq. km. Then you look at New York City, which has about 10,600 people per sq. km. Not to mention the fact that in New York, 1% of the population controls 1/3 of the wealth (according to adjusted federal taxable income). I don't know what Helsinki's economic disparity is, but Finland's national economic disparity is about a third of what it is for the US.
Maybe we shouldn't be asking "how do we fix education" but rather "how do we curb the economic factors that hurt education".
Basically I'm just reiterating what was brought up on the first page:
The education problem in this country is actually a poverty problem. Full stop.
The issue isn't even school funding per se, although that is a confounding factor in many districts. The problem is that children in poor neighbourhoods do not come to school prepared to learn. The achievement gap between rich and poor and black and white is evident before formal schooling even begins. This is the result of a constellation of problems that plague poor children. These problems include but are not limited to lack of intellectual stimulation in the home during critical periods of brain development, malnutrition, health problems resulting from lack of access to care, health problems resulting from exposure to environmental toxins, and parents who are themselves poorly educated and do not know how to support their children's learning at home.
Research does show that teacher quality is the most important factor in a child's education within the school itself. However, socioeconomic status is overwhelmingly more important to a child's success in school than anything that happens in the classroom.
One big reason people look at Finland is that it wasn't always that great. Finland didn't slowly drift up, and it's not a case of a terrible system just gaining competency and modernity. Finland went from sub-par to the best in, what, thirty years?
OR apply it to cops. Should we grade cops by how many arrests they make? That's their job right?
Except some students have professor and scientist parents areas are already fairly crime free. So do you grade this cop down because he's not making improvements arresting people? Or do you grade him higher than the cop in the inner city because improving from the bottom is easiermore crimes go unsolved there? Is a result or change of results arrest or change in crime the same everywhere?
You end up spending so much time money and energy not only trying to come up with a reasonable metric and evaluation regime but figuring out what to do with those measurements that you end up in a huge hole before you even start making improvements.
The saddest part of this is that in many areas the bolded is literally true. Quotas and all sorts of ways police organizations rate the performance of their officers are based around the number of people they can catch and it's thanks to this we have policies like "Stop and Frisk" which are basically about racking these numbers up further. and it is literally damaging the entire social fabric by sowing a deep distrust of Law Enforcement. This might not be the best analogy to use for that reason though, since there are a lot of people that still get very very touchy the moment you mention that the majority of cops are not upstanding moral paragons.
If the best you can come up with against someone who's patently ignorant is to yell back at him, "Yeah? Well there's BOOKS, and they say you're WRONG!"
Then honestly you're not coming out of this looking great either.
I ran into an interesting article on school spending vs school quality broken down to a district level. It talks about the lack of correlation between spending and performance, but it also brings up mention of specific spending issues that lend some support for what I was wondering earlier.
In particular, he found that urban schools tended to spend inefficiently for a variety of reasons, including high staff and student turnover and conflicts over how to teach struggling students. At the same time, he said, urban districts often have extra expenses for needs such as security, dropout prevention, or for teaching students who are not proficient in English.
Deanne Fitzmaurice/California WatchTeacher Mellissa Schmitz high-fives a student at Costano Elementary School in Ravenswood City School District, East Palo Alto.
One of those districts with higher expenses is the Ravenswood City School District in East Palo Alto, where about two-thirds of students are English learners. The district has had to hire three full-time Spanish translators – mainly to translate lengthy special education reports as required by law – and has translators working in the school office, in classrooms and at parent meetings.
Ravenswood spends nearly $13,000 per student, yet has cut several programs and may slash two weeks from the next school year, said Superintendent Maria De La Vega.
The correlation isn't perfect, and there are a few counter-instances involving frugal and efficient use of dollars, but what I'm getting from the article is that a lot of the issue is that schools are not getting the funding they might need on a per-school basis to address their specific needs. And that, specifically, you can't just look at the final per-student spending for a school as if that is the final word on whether that school is adequately funded.
There's really not much of an issue in my area with ESL kids, so I don't know how they're handled, but my honest belief is that in areas of exceptionally high ESL density you should run a second school for ESL kids that does literally nothing but get them proficient in English as quickly as possible. Nothing in this country is going to fuck you as hard as speaking English poorly.
OR apply it to cops. Should we grade cops by how many arrests they make? That's their job right?
Except some students have professor and scientist parents areas are already fairly crime free. So do you grade this cop down because he's not making improvements arresting people? Or do you grade him higher than the cop in the inner city because improving from the bottom is easiermore crimes go unsolved there? Is a result or change of results arrest or change in crime the same everywhere?
You end up spending so much time money and energy not only trying to come up with a reasonable metric and evaluation regime but figuring out what to do with those measurements that you end up in a huge hole before you even start making improvements.
The saddest part of this is that in many areas the bolded is literally true. Quotas and all sorts of ways police organizations rate the performance of their officers are based around the number of people they can catch and it's thanks to this we have policies like "Stop and Frisk" which are basically about racking these numbers up further. and it is literally damaging the entire social fabric by sowing a deep distrust of Law Enforcement. This might not be the best analogy to use for that reason though, since there are a lot of people that still get very very touchy the moment you mention that the majority of cops are not upstanding moral paragons.
I once had a conversation with a cop where she was complaining because where she normally was on patrol at night there weren't that many drunk drivers, and she sucked at catching them, so it was hurting her performance evaluations. She even said she'd pull over pretty much anyone at 2AM and basically invent a reason to do it.
Every modern country seems to have this kind of endless debate, if it is any consolation. NZ at this very moment is bringing in league tables and charter schools for no obvious reason bar ideology (as our public system is usually thought to work well enough).
It's really worth reading Mother Jones Magazine's piece on Mission High in San Francisco. Objectively, it's doing everything we want a school to do - they're working with at-risk students to get them looking upward. But according to the results of the STAR exam, they're considered a failing school. Something is wrong here, and I don't think it's the school.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
Specifically? I don't think anybody in here is qualified to design an evaluation system in specific terms, because I don't think any of us are educators. You may as well ask us to specifically map out the specs for a proper manned rocket to Mars.
Things that should play into it, in broad terms:
- How are the kids performing at the end of the year?
- How were they performing at the beginning of the year?
- What sort of problem incidents occured during the year, and how did the teacher respond?
- How does the teacher engage and interact with the students?
- How does the teacher engage and interact with the parents?
- How has the teacher dealt with the limitations she's run across in her duties?
Then these things can be weighed against the realities of the classroom, including how prepared the children are to learn, how many special needs kids exist in the class, the quality of the teaching materials available, and so on.
And as Tenek said, it would be fuzzy. If your evaluation system can boil everything down into a single letter grade, it is probably a shitty evaluation system.
I submitted an entry to Lego Ideas, and if 10,000 people support me, it'll be turned into an actual Lego set!If you'd like to see and support my submission, follow this link.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
At a very high level?
Determine the techniques and procedures you want your teachers to use, based on available empirical research. I'm being deliberately vague here because I'm not well-versed in educational models.
Fire the teachers who consistently fail to apply those techniques and procedures.
Put particular scrutiny on teachers whose students perform particularly worse than would be expected from their demographic.
This makes the occupation procedure-oriented, rather than results-oriented.
Why do it this way? Because if you look at other public service employees who have to work with a diverse, sometimes hostile, population, where results are more the product of the "customer"'s social environment, that's how it works. Police officers, physicians, social workers. Ideally, they're evaluated primarily by how well they conform with established guidelines, and those guidelines are (again, ideally) established by looking at trends.
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
Put particular scrutiny on teachers whose students perform particularly worse than would be expected from their demographic.
This is key. If the kids from Beverly Hills Unified are even in the same spreadsheet as the kids from East Compton Unified, you're doing something very wrong.
I think the comparison to how we might grade physicians is a good one, and the current mindset in education seems sort of like comparing the survivial rates of oncology patients versus plastic surgery patients to see which ones had better doctors.
It's like organic orchard-grown apples versus oranges that grew in a closet somewhere in the vapor cloud of some dude's meth lab and also the oranges don't speak english.
I submitted an entry to Lego Ideas, and if 10,000 people support me, it'll be turned into an actual Lego set!If you'd like to see and support my submission, follow this link.
BTW you see a similar effect, though for different reasons, in higher education - college costs are going up and they're largely being driven by administrative, non-teaching-staff, and facilities expenditures.
While that's part of it, I would contend that a larger part of what's driving up college costs are the continual cuts in state funding.
Yup. It's a pretty easy correlation to prove - I've seen the graphs. Salaries - even the much touted administrator salaries - only account for a fraction of the higher tuition. The lion share comes from state legislators having substantially cut funding to public colleges. In some states, the share made up by state appropriations is less than a quarter of what it was in the 1980s.
Not really the same thing.
When I say 'college costs' I'm talking about total operational costs. I'm not talking about tuition prices.
You guys are talking about who foots the bill. Does the state foot the bill through subsidies, or do students foot the bill through loans?
That's a good discussion to have, but the deeper underlying question is - why has the bill gotten so big? You can't answer that question by looking at funding sources.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
Scholarships have also been going up, which is likely a good thing.
Yes and no.
It's good that scholarships are going up, but this is another reflection of the rising burden on the students and families. A lot of scholarships are funded by tuition, so the burden is just spread out a bit.
And that's also coming under attack. There's a strong move on in conservative circles to curtail or outright ban using tuition monies to pay for scholarships. They say its not fair to force families to subsidize poor students. This is definitely a bubbling talking point, originating from think tanks like the Civitas Institute.
You can get around that by shuffling like 4 lines on a financial statement. Seems like a stupid thing to go after.
In most public schools, a certain amount of scholarships are cooked into everyone's base tuition. Every time tuition has gone up, the universities here have made a point of noting that some percent (usually 15-25 percent) of the increase will go toward need-based scholarships. That's gotten the conservatives in a dander, since its government redistributing their wealth.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
Handwritten stuff poses a bit of an issue, it takes years to generate results, and requires infrastructure, but bringing business analytics to bear one the problem could generate massive amounts of objective information about education processes. You would be able to correct for things like demographic disparities as well.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
Handwritten stuff poses a bit of an issue, it takes years to generate results, and requires infrastructure, but bringing business analytics to bear one the problem could generate massive amounts of objective information about education processes. You would be able to correct for things like demographic disparities as well.
If you can build a computer like that you can just fire all the teachers and have it teach the kids directly. You have blithely dumped the next fifty years of AI research into one sentence like it's already done and you just need to bang on the keyboard for a few seconds. This stuff is hard for humans to measure, never mind machines.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
Handwritten stuff poses a bit of an issue, it takes years to generate results, and requires infrastructure, but bringing business analytics to bear one the problem could generate massive amounts of objective information about education processes. You would be able to correct for things like demographic disparities as well.
If you can build a computer like that you can just fire all the teachers and have it teach the kids directly. You have blithely dumped the next fifty years of AI research into one sentence like it's already done and you just need to bang on the keyboard for a few seconds. This stuff is hard for humans to measure, never mind machines.
Not AI, statistics.
You don't think a computer can look at a math problem and figure out the sort of math needed to solve it, and see if it was solved correctly? You don't think a computer can see how long words used in an essay are and how long and properly structured sentences are?
History? Well, you could get wolfram alpha to spit out a digest based on the covered materials, and then run something like Watson over it to see which of those things are covered in the essay,
but it would probably be easier just to have the teacher grade them(which they kinda would have to do anyway), and then look grade given to the kids by their other teacher that cover the similar material. Probably weight based on a sigma squares dealie, like you would with any other loosely coupled discrete bunch of data points
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
This was actually in the MJ article that AngelHedgie linked.
The problem? Still relying almost entirely on standardized testing. No matter how awesomely sensitive your regression analysis is, you're still giving teachers incentive to teach to the Scantron.
every person who doesn't like an acquired taste always seems to think everyone who likes it is faking it. it should be an official fallacy.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
Handwritten stuff poses a bit of an issue, it takes years to generate results, and requires infrastructure, but bringing business analytics to bear one the problem could generate massive amounts of objective information about education processes. You would be able to correct for things like demographic disparities as well.
If you can build a computer like that you can just fire all the teachers and have it teach the kids directly. You have blithely dumped the next fifty years of AI research into one sentence like it's already done and you just need to bang on the keyboard for a few seconds. This stuff is hard for humans to measure, never mind machines.
Not AI, statistics.
You don't think a computer can look at a math problem and figure out the sort of math needed to solve it, and see if it was solved correctly? You don't think a computer can see how long words used in an essay are and how long and properly structured sentences are?
History? Well, you could get wolfram alpha to spit out a digest based on the covered materials, and then run something like Watson over it to see which of those things are covered in the essay,
but it would probably be easier just to have the teacher grade them(which they kinda would have to do anyway), and then look grade given to the kids by their other teacher that cover the similar material. Probably weight based on a sigma squares dealie, like you would with any other loosely coupled discrete bunch of data points
Yeah, no. Some things defy quantization by their very nature, and this is one of them. Which is the whole underlying issue here.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
This was actually in the MJ article that AngelHedgie linked.
The problem? Still relying almost entirely on standardized testing. No matter how awesomely sensitive your regression analysis is, you're still giving teachers incentive to teach to the Scantron.
I am basically adding an extra year, talking about assignments rather than standardized tests, and weighting by how consitently the teachers grade students(which breaks a little with small schools or narrow tracks where trigger, pre calc, and calc are all taught by the same teachers and every student takes them). The result basically is teachers are graded by how well their kid do in the next class, compared to the class before, while trying to control for the contributions of those two other teachers.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
This was actually in the MJ article that AngelHedgie linked.
The problem? Still relying almost entirely on standardized testing. No matter how awesomely sensitive your regression analysis is, you're still giving teachers incentive to teach to the Scantron.
I am basically adding an extra year, talking about assignments rather than standardized tests, and weighting by how consitently the teachers grade students(which breaks a little with small schools or narrow tracks where trigger, pre calc, and calc are all taught by the same teachers and every student takes them). The result basically is teachers are graded by how well their kid do in the next class, compared to the class before, while trying to control for the contributions of those two other teachers.
This is a really good example of the underlying problems of assessing teachers. That formula would produce a number that is understandable by anyone. This teacher is 9.1, this one is only 3.6. And then we know who is better.
But that's ridiculous. I can't assess this job, and I shouldn't be either. Their bosses and peers can assess them, using criteria such a El Jeffe posted above. We outsiders can't, any more than I could possibly assess if Spool does his job well.
We want dumb metrics so we can all judge teachers. Why do we want simple and useless metrics? Because we 'know' teaching is something we are all qualified to judge, unlike any other skilled profession involving people and society.
Look, a salesman's success can be measured in how many dollars in sales they have.
But with a teacher you're measuring what? How much the child learned? Setting aside the fact that the teacher doesn't have total control of that, what's the scale of measurement? Do you measure each use of a concept that is at or above that grade level as one Learntron. If your students accumulate enough Learntrons, congratulations you're a great teacher. I'm not even sure any more what I learned in each grade, and it's not because I'm just getting old. I learned the same facts numerous times. When I was a kid, it would have been silly to get me to sit down and tell someone what I learned, or demonstrate only what I learned that year. I wouldn't even know necessarily all of the things that I should cover.
Another problem is that when I was about seven years old, I memorized every state capitol, and I would get my parents to quiz me. I can still, with a little thought, probably give them all back to you. But I taught myself that. And I taught it to myself about four years or so before I was required to know it for a class. By the time I got to the class it was easy. Should my teacher for that grade get credit because I learned something a long time ago? That might result in teachers that aren't very good getting by because while they have a bunch of mediocre students, they have a few really great ones. Turns out those really great ones were already that way.
"The only way to get rid of a temptation is to give into it." - Oscar Wilde
"We believe in the people and their 'wisdom' as if there was some special secret entrance to knowledge that barred to anyone who had ever learned anything." - Friedrich Nietzsche
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
This was actually in the MJ article that AngelHedgie linked.
The problem? Still relying almost entirely on standardized testing. No matter how awesomely sensitive your regression analysis is, you're still giving teachers incentive to teach to the Scantron.
I am basically adding an extra year, talking about assignments rather than standardized tests, and weighting by how consitently the teachers grade students(which breaks a little with small schools or narrow tracks where trigger, pre calc, and calc are all taught by the same teachers and every student takes them). The result basically is teachers are graded by how well their kid do in the next class, compared to the class before, while trying to control for the contributions of those two other teachers.
This is a really good example of the underlying problems of assessing teachers. That formula would produce a number that is understandable by anyone. This teacher is 9.1, this one is only 3.6. And then we know who is better.
But that's ridiculous. I can't assess this job, and I shouldn't be either. Their bosses and peers can assess them, using criteria such a El Jeffe posted above. We outsiders can't, any more than I could possibly assess if Spool does his job well.
We want dumb metrics so we can all judge teachers. Why do we want simple and useless metrics? Because we 'know' teaching is something we are all qualified to judge, unlike any other skilled profession involving people and society.
and there is no point in having ways to compare different teaching styles, processes and methods between systems and schools? Which is kinda something I care about more than deciding what kind of raise a teacher should receive.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
This was actually in the MJ article that AngelHedgie linked.
The problem? Still relying almost entirely on standardized testing. No matter how awesomely sensitive your regression analysis is, you're still giving teachers incentive to teach to the Scantron.
I am basically adding an extra year, talking about assignments rather than standardized tests, and weighting by how consitently the teachers grade students(which breaks a little with small schools or narrow tracks where trigger, pre calc, and calc are all taught by the same teachers and every student takes them). The result basically is teachers are graded by how well their kid do in the next class, compared to the class before, while trying to control for the contributions of those two other teachers.
This would become a massive data entry problem. Unless all class work assignments are standardized each teacher is going to have to categorize the answers from every assignment. Then after grading they would have to enter the score for each category. Depending on the subject the categorization could be subjective and the grading of the answer could be subjective. If the teacher administering the assignment is also in charge of data entry / categorization then they have an incentive to give students a higher grade so they don't appear like a bad teacher according to the algorithm.
I'm not saying this couldn't be done but I'm doubting teachers would have the time to do this extra sort of work so that would require a lot more man power.
CommunistCow on
No, I am not really communist. Yes, it is weird that I use this name.
and there is no point in having ways to compare different teaching styles, processes and methods between systems and schools? Which is kinda something I care about more than deciding what kind of raise a teacher should receive.
Sure there is. And there have been studies that attempt to weigh the merits of various teaching styles and processes and methods. But not a single one of them has sought to boil the whole comparison down to a single number. And they always come with a bunch of caveats and qualifiers pertaining to what kind of students we're talking about, what grade levels, and what very specific skills are being measured.
The short answer is that there isn't a "best" method. The long answer is basically the entirety of several thousand pages of research papers.
I submitted an entry to Lego Ideas, and if 10,000 people support me, it'll be turned into an actual Lego set!If you'd like to see and support my submission, follow this link.
Naw, not really. That would add precision(and eventually the computer being able to figure it out by itself would be nice, and I think we are close to that for most subjects that folk might expect), but wouldn't be necessary. You can do pretty much the same math just knowing how much of what skills the assignment consisted of, approximately. Teachers have to create fairly detailed lesson plans already, it not be horrific to expect them to show they are dealing with specific skills based on specific assignments. Shows they have aren't assigning busy work and are meeting the goals for cover skills. It would be handful of criteria per assignment, that's pretty trivial compared to the time taken thinking them up, figuring out how to issue instructions, grading them, and recording the grades. It's entering skills into a computer or assigning weights to a list of skills they created earlier. That's adding, what? 2 minutes to a process that probably takes several hours? Yeah, that is mountains of extra work.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
Perhaps because we don't accept there is a hard and fast way of doing so?
I'm a programmer. There isn't a good metric for evaluating my performance. You can count tasks, count bug fixes, count lines of code but those are all poor methods for evaluation. A company that did so would get poor results.
It also assumes that the problem is bad teachers or that punishing bad teachers (or rewarding good ones) is central to the problem. But even if you fire the worst 10% of teachers, you'll just end up with bigger classes and worse teachers replacing them. There isn't supply of high quality teachers being denied a chance by bad teachers.
and there is no point in having ways to compare different teaching styles, processes and methods between systems and schools? Which is kinda something I care about more than deciding what kind of raise a teacher should receive.
Sure there is. And there have been studies that attempt to weigh the merits of various teaching styles and processes and methods. But not a single one of them has sought to boil the whole comparison down to a single number. And they always come with a bunch of caveats and qualifiers pertaining to what kind of students we're talking about, what grade levels, and what very specific skills are being measured.
The short answer is that there isn't a "best" method. The long answer is basically the entirety of several thousand pages of research papers.
So there needs to be some kind of organization that collects the research results and creates a best fit method that accounts for all the differences in student needs. Just because one can't say that method A works for all students shouldn't preclude us from using it if it works well for inner city black 4th graders of single mothers while method B works for rural Hispanic 6th graders with learning disabilities.
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I think the best way to evaluate teachers is for them to be given specific methods and tasks that have been proven by research to be effective and to observe if and how they implement the methods. If you create clear, measurable behavioral goals for teachers we can ensure that the classroom experience for all students meet the expectations that research and analysis have shown to be optimal for student achievement. Developing what those actions the teachers have to take, from evaluating student needs to creating individual educational plans to whatever, should be one focus. The other should be creating a management style in schools that focuses on improving teacher behavior to the goals set out.
This way if a teacher fails despite implementing the programs correctly, we can see it as a failure of the program in this case and we can go back and develop a new one.
He's a shy overambitious dog-catcher on the wrong side of the law. She's an orphaned psychic mercenary with the power to bend men's minds. They fight crime!
Again and again, we've talked about ways we should not evaluate teachers.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
This was actually in the MJ article that AngelHedgie linked.
The problem? Still relying almost entirely on standardized testing. No matter how awesomely sensitive your regression analysis is, you're still giving teachers incentive to teach to the Scantron.
Union teacher here. We're fighting the shit out of this as we speak. It's insane. My state government just lost in lower court because their own experts could not explain what the hell that equation measures to the judge.
Want to evaluate teachers? Peer evaluations and administrative evaluations based on actual time spent in the teacher's classroom, watching lessons and interacting with students. Base the evaluations on current research in pedagogy. Throw the tests out the window.
The numbers are goosery. Kids arent RPG characters; they don't gain 47 EXP in Critical Thinking and 62 EXP in Systems of Equations.
Naw, not really. That would add precision(and eventually the computer being able to figure it out by itself would be nice, and I think we are close to that for most subjects that folk might expect), but wouldn't be necessary. You can do pretty much the same math just knowing how much of what skills the assignment consisted of, approximately. Teachers have to create fairly detailed lesson plans already, it not be horrific to expect them to show they are dealing with specific skills based on specific assignments. Shows they have aren't assigning busy work and are meeting the goals for cover skills. It would be handful of criteria per assignment, that's pretty trivial compared to the time taken thinking them up, figuring out how to issue instructions, grading them, and recording the grades. It's entering skills into a computer or assigning weights to a list of skills they created earlier. That's adding, what? 2 minutes to a process that probably takes several hours? Yeah, that is mountains of extra work.
You are not understanding how much work is involved in assessing a student. Things that are easy for humans to do are not automatically easy for computers. Making a detailed lesson plan means that another person who reads it over can understand it. The machine is going to be completely clueless (but it might do real well at OCR and data storage). All you can do right now is take whatever grades the human assigns to each piece of work and run some stats on them. At best what you end up with is a performance metric that rewards teachers for teaching to the test. Test scores stop being a good evaluation of a student's performance the moment you use them to evaluate the teacher.
Posts
Not really the same thing.
When I say 'college costs' I'm talking about total operational costs. I'm not talking about tuition prices.
You guys are talking about who foots the bill. Does the state foot the bill through subsidies, or do students foot the bill through loans?
That's a good discussion to have, but the deeper underlying question is - why has the bill gotten so big? You can't answer that question by looking at funding sources.
the "no true scotch man" fallacy.
The correlation isn't perfect, and there are a few counter-instances involving frugal and efficient use of dollars, but what I'm getting from the article is that a lot of the issue is that schools are not getting the funding they might need on a per-school basis to address their specific needs. And that, specifically, you can't just look at the final per-student spending for a school as if that is the final word on whether that school is adequately funded.
And I'd counter that the bill hasn't actually gotten that much bigger. In fact, the big investments - infrastructure costs that were paid for by the state to build the schools in the first place - have gone down. Overall funding, though, hasn't climbed nearly to the levels that the people who demonize higher education want you to believe.
Basically, the costs haven't skyrocketed. The funding sources have just shifted from the taxpayers to the individual students. The glory days of higher education spending were actually in the 1940s/1950s (post-GI bill) and 1970s (boomer expansion). We've held steady since.
http://www.usgovernmentspending.com/downchart_gs.php?year=1900_2014&units=p&chart=20-fed_20-statelocal&title=Total Education Spending
Like a lot of American funding crises, the real problem has to do with the fact that we've been underfunding society since in the 1980s, and it's finally catching up with us. You can't cut taxes forever and expect to remain a First World nation.
Scholarships have also been going up, which is likely a good thing.
Yes and no.
It's good that scholarships are going up, but this is another reflection of the rising burden on the students and families. A lot of scholarships are funded by tuition, so the burden is just spread out a bit.
And that's also coming under attack. There's a strong move on in conservative circles to curtail or outright ban using tuition monies to pay for scholarships. They say its not fair to force families to subsidize poor students. This is definitely a bubbling talking point, originating from think tanks like the Civitas Institute.
I don't know why you're linking a chart on US government spending when that is largely irrelevant to what I'm talking about.
the "no true scotch man" fallacy.
Quick Googling. I need to find better data, but I've been to enough presentations on the subject to know that the basic idea is the correct - college costs aren't skyrocketing anywhere near to the point that the political rhetoric makes out. It's just that students and parents are shouldering more of the burden.
There's fat to be cut in administration - although a lot of that administrative growth has arisen from the need to develop and maintain additional funding sources. There's fat to be cut in dorms and on-campus dining - although cutting back runs into free market problems in that schools have to maintain housing and food choices but anything below a certain level of quality will not be competitive with the private sector, students will eat/live elsewhere and the lack of profits will end up costing schools more money to maintain.
The issue is actually pretty simple. States have cut, on average, 75 percent of their funding of public schools. The federal government is fickle as well, cutting back grants in favor of loans. Colleges are expensive to maintain and that money has to come from somewhere.
But the reality of this conversation has been drowned out in the bullshit.
You can get around that by shuffling like 4 lines on a financial statement. Seems like a stupid thing to go after.
In most public schools, a certain amount of scholarships are cooked into everyone's base tuition. Every time tuition has gone up, the universities here have made a point of noting that some percent (usually 15-25 percent) of the increase will go toward need-based scholarships. That's gotten the conservatives in a dander, since its government redistributing their wealth.
OR apply it to cops. Should we grade cops by how many arrests they make? That's their job right?
Except some students have professor and scientist parents areas are already fairly crime free. So do you grade this cop down because he's not making improvements arresting people? Or do you grade him higher than the cop in the inner city because improving from the bottom is easiermore crimes go unsolved there? Is a result or change of results arrest or change in crime the same everywhere?
You end up spending so much time money and energy not only trying to come up with a reasonable metric and evaluation regime but figuring out what to do with those measurements that you end up in a huge hole before you even start making improvements.
QEDMF xbl: PantsB G+
Maybe the dollars it takes to educate kids goes up exponentially rather than linearly as population density and economic disparity increases. Helsinki, I believe, is Finland's most densely populated city with a whopping 2700 people per sq. km. Then you look at New York City, which has about 10,600 people per sq. km. Not to mention the fact that in New York, 1% of the population controls 1/3 of the wealth (according to adjusted federal taxable income). I don't know what Helsinki's economic disparity is, but Finland's national economic disparity is about a third of what it is for the US.
Maybe we shouldn't be asking "how do we fix education" but rather "how do we curb the economic factors that hurt education".
Basically I'm just reiterating what was brought up on the first page:
Total education costs per student per year have been increasing faster than inflation or CPI since 1970. This is a noncontroversial statement. It's also not a new trend - it's a 40-year trend that has only recently come into the public eye due to the withdrawal of public funding sources. The public and the media didn't care nearly as much about the rising cost of education as long as the states and the feds were footing the bill.
So I guess if you're arguing against the idea that there's been a sudden 10-year spike in education costs... well, I'd agree with you there. That's been mostly driven by the withdrawal in state funding.
But I think the evidence is clear that there's also a longer-term, slower trend.
As for the need to develop and maintain additional funding sources, I agree that this is a problem, but I am skeptical that this is a majority cost driver. The book Tuition Rising breaks down in excruciating detail the expenditures of multiple institutions, and it shows that the need to hire more people to manage increasingly complex regulations (to secure funding, among other things) is a driver of personnel costs but it's not a majority one.
And I recognize the CCAP is a right-leaning organization, but the data behind their "25 Ways to Reduce the Cost of College" white paper (PDF summary) is hard to dismiss. (Their conclusions are also mirrored in Tuition Rising.) Both teaching and nonteaching payroll have been increasing (faster than inflation), and the latter even faster than the former. Some even clearer explanations are given by one of the CCAP fellows here: http://www.mindingthecampus.com/originals/2010/03/why_the_student_protestors_are.html and here: http://www.forbes.com/forbes/2009/0713/opinions-college-tuition-teachers-on-my-mind.html
Here's the problem - simply restoring public funding to pre-1999 (or whatever recent year you choose) levels just papers over the problem. There are perverse economic incentives at play that stimulate spending on non-educational expenses and deincentivize cost cutting. Increasing student aid just pours more money into a cracked vessel.
the "no true scotch man" fallacy.
One big reason people look at Finland is that it wasn't always that great. Finland didn't slowly drift up, and it's not a case of a terrible system just gaining competency and modernity. Finland went from sub-par to the best in, what, thirty years?
The saddest part of this is that in many areas the bolded is literally true. Quotas and all sorts of ways police organizations rate the performance of their officers are based around the number of people they can catch and it's thanks to this we have policies like "Stop and Frisk" which are basically about racking these numbers up further. and it is literally damaging the entire social fabric by sowing a deep distrust of Law Enforcement. This might not be the best analogy to use for that reason though, since there are a lot of people that still get very very touchy the moment you mention that the majority of cops are not upstanding moral paragons.
There's really not much of an issue in my area with ESL kids, so I don't know how they're handled, but my honest belief is that in areas of exceptionally high ESL density you should run a second school for ESL kids that does literally nothing but get them proficient in English as quickly as possible. Nothing in this country is going to fuck you as hard as speaking English poorly.
I once had a conversation with a cop where she was complaining because where she normally was on patrol at night there weren't that many drunk drivers, and she sucked at catching them, so it was hurting her performance evaluations. She even said she'd pull over pretty much anyone at 2AM and basically invent a reason to do it.
How should we evaluate teachers? Specifically, not in broad terms. If you were designing an evaluation system, what would it look like?
It would be fuzzy and partially reliant on the judgment of other teachers. Goodhart's law is going to rip any overly-specific plan to shreds.
Specifically? I don't think anybody in here is qualified to design an evaluation system in specific terms, because I don't think any of us are educators. You may as well ask us to specifically map out the specs for a proper manned rocket to Mars.
Things that should play into it, in broad terms:
- How are the kids performing at the end of the year?
- How were they performing at the beginning of the year?
- What sort of problem incidents occured during the year, and how did the teacher respond?
- How does the teacher engage and interact with the students?
- How does the teacher engage and interact with the parents?
- How has the teacher dealt with the limitations she's run across in her duties?
Then these things can be weighed against the realities of the classroom, including how prepared the children are to learn, how many special needs kids exist in the class, the quality of the teaching materials available, and so on.
And as Tenek said, it would be fuzzy. If your evaluation system can boil everything down into a single letter grade, it is probably a shitty evaluation system.
At a very high level?
Determine the techniques and procedures you want your teachers to use, based on available empirical research. I'm being deliberately vague here because I'm not well-versed in educational models.
Fire the teachers who consistently fail to apply those techniques and procedures.
Put particular scrutiny on teachers whose students perform particularly worse than would be expected from their demographic.
This makes the occupation procedure-oriented, rather than results-oriented.
Why do it this way? Because if you look at other public service employees who have to work with a diverse, sometimes hostile, population, where results are more the product of the "customer"'s social environment, that's how it works. Police officers, physicians, social workers. Ideally, they're evaluated primarily by how well they conform with established guidelines, and those guidelines are (again, ideally) established by looking at trends.
the "no true scotch man" fallacy.
This is key. If the kids from Beverly Hills Unified are even in the same spreadsheet as the kids from East Compton Unified, you're doing something very wrong.
I think the comparison to how we might grade physicians is a good one, and the current mindset in education seems sort of like comparing the survivial rates of oncology patients versus plastic surgery patients to see which ones had better doctors.
It's like organic orchard-grown apples versus oranges that grew in a closet somewhere in the vapor cloud of some dude's meth lab and also the oranges don't speak english.
I really believe you could learn a lot about how effective educators by dumping all their students' work into a big ass computer, defining success criteria(more or less th ability of student to display whatever skills they are supposed to be teaching) and looking at how well those students displayed those skills the year before, the year the instructor taught them, and the subsequent year. It would give you a decent delta on how the teacher performed, how they thought they were performing, and probably what sort of students benefit most from their teaching methods.
Handwritten stuff poses a bit of an issue, it takes years to generate results, and requires infrastructure, but bringing business analytics to bear one the problem could generate massive amounts of objective information about education processes. You would be able to correct for things like demographic disparities as well.
If you can build a computer like that you can just fire all the teachers and have it teach the kids directly. You have blithely dumped the next fifty years of AI research into one sentence like it's already done and you just need to bang on the keyboard for a few seconds. This stuff is hard for humans to measure, never mind machines.
Not AI, statistics.
You don't think a computer can look at a math problem and figure out the sort of math needed to solve it, and see if it was solved correctly? You don't think a computer can see how long words used in an essay are and how long and properly structured sentences are?
History? Well, you could get wolfram alpha to spit out a digest based on the covered materials, and then run something like Watson over it to see which of those things are covered in the essay,
but it would probably be easier just to have the teacher grade them(which they kinda would have to do anyway), and then look grade given to the kids by their other teacher that cover the similar material. Probably weight based on a sigma squares dealie, like you would with any other loosely coupled discrete bunch of data points
This was actually in the MJ article that AngelHedgie linked.
The problem? Still relying almost entirely on standardized testing. No matter how awesomely sensitive your regression analysis is, you're still giving teachers incentive to teach to the Scantron.
the "no true scotch man" fallacy.
Yeah, no. Some things defy quantization by their very nature, and this is one of them. Which is the whole underlying issue here.
I am basically adding an extra year, talking about assignments rather than standardized tests, and weighting by how consitently the teachers grade students(which breaks a little with small schools or narrow tracks where trigger, pre calc, and calc are all taught by the same teachers and every student takes them). The result basically is teachers are graded by how well their kid do in the next class, compared to the class before, while trying to control for the contributions of those two other teachers.
This is a really good example of the underlying problems of assessing teachers. That formula would produce a number that is understandable by anyone. This teacher is 9.1, this one is only 3.6. And then we know who is better.
But that's ridiculous. I can't assess this job, and I shouldn't be either. Their bosses and peers can assess them, using criteria such a El Jeffe posted above. We outsiders can't, any more than I could possibly assess if Spool does his job well.
We want dumb metrics so we can all judge teachers. Why do we want simple and useless metrics? Because we 'know' teaching is something we are all qualified to judge, unlike any other skilled profession involving people and society.
But with a teacher you're measuring what? How much the child learned? Setting aside the fact that the teacher doesn't have total control of that, what's the scale of measurement? Do you measure each use of a concept that is at or above that grade level as one Learntron. If your students accumulate enough Learntrons, congratulations you're a great teacher. I'm not even sure any more what I learned in each grade, and it's not because I'm just getting old. I learned the same facts numerous times. When I was a kid, it would have been silly to get me to sit down and tell someone what I learned, or demonstrate only what I learned that year. I wouldn't even know necessarily all of the things that I should cover.
Another problem is that when I was about seven years old, I memorized every state capitol, and I would get my parents to quiz me. I can still, with a little thought, probably give them all back to you. But I taught myself that. And I taught it to myself about four years or so before I was required to know it for a class. By the time I got to the class it was easy. Should my teacher for that grade get credit because I learned something a long time ago? That might result in teachers that aren't very good getting by because while they have a bunch of mediocre students, they have a few really great ones. Turns out those really great ones were already that way.
"We believe in the people and their 'wisdom' as if there was some special secret entrance to knowledge that barred to anyone who had ever learned anything." - Friedrich Nietzsche
and there is no point in having ways to compare different teaching styles, processes and methods between systems and schools? Which is kinda something I care about more than deciding what kind of raise a teacher should receive.
This would become a massive data entry problem. Unless all class work assignments are standardized each teacher is going to have to categorize the answers from every assignment. Then after grading they would have to enter the score for each category. Depending on the subject the categorization could be subjective and the grading of the answer could be subjective. If the teacher administering the assignment is also in charge of data entry / categorization then they have an incentive to give students a higher grade so they don't appear like a bad teacher according to the algorithm.
I'm not saying this couldn't be done but I'm doubting teachers would have the time to do this extra sort of work so that would require a lot more man power.
Sure there is. And there have been studies that attempt to weigh the merits of various teaching styles and processes and methods. But not a single one of them has sought to boil the whole comparison down to a single number. And they always come with a bunch of caveats and qualifiers pertaining to what kind of students we're talking about, what grade levels, and what very specific skills are being measured.
The short answer is that there isn't a "best" method. The long answer is basically the entirety of several thousand pages of research papers.
Perhaps because we don't accept there is a hard and fast way of doing so?
I'm a programmer. There isn't a good metric for evaluating my performance. You can count tasks, count bug fixes, count lines of code but those are all poor methods for evaluation. A company that did so would get poor results.
It also assumes that the problem is bad teachers or that punishing bad teachers (or rewarding good ones) is central to the problem. But even if you fire the worst 10% of teachers, you'll just end up with bigger classes and worse teachers replacing them. There isn't supply of high quality teachers being denied a chance by bad teachers.
QEDMF xbl: PantsB G+
So there needs to be some kind of organization that collects the research results and creates a best fit method that accounts for all the differences in student needs. Just because one can't say that method A works for all students shouldn't preclude us from using it if it works well for inner city black 4th graders of single mothers while method B works for rural Hispanic 6th graders with learning disabilities.
I think the best way to evaluate teachers is for them to be given specific methods and tasks that have been proven by research to be effective and to observe if and how they implement the methods. If you create clear, measurable behavioral goals for teachers we can ensure that the classroom experience for all students meet the expectations that research and analysis have shown to be optimal for student achievement. Developing what those actions the teachers have to take, from evaluating student needs to creating individual educational plans to whatever, should be one focus. The other should be creating a management style in schools that focuses on improving teacher behavior to the goals set out.
This way if a teacher fails despite implementing the programs correctly, we can see it as a failure of the program in this case and we can go back and develop a new one.
Union teacher here. We're fighting the shit out of this as we speak. It's insane. My state government just lost in lower court because their own experts could not explain what the hell that equation measures to the judge.
Want to evaluate teachers? Peer evaluations and administrative evaluations based on actual time spent in the teacher's classroom, watching lessons and interacting with students. Base the evaluations on current research in pedagogy. Throw the tests out the window.
The numbers are goosery. Kids arent RPG characters; they don't gain 47 EXP in Critical Thinking and 62 EXP in Systems of Equations.
You are not understanding how much work is involved in assessing a student. Things that are easy for humans to do are not automatically easy for computers. Making a detailed lesson plan means that another person who reads it over can understand it. The machine is going to be completely clueless (but it might do real well at OCR and data storage). All you can do right now is take whatever grades the human assigns to each piece of work and run some stats on them. At best what you end up with is a performance metric that rewards teachers for teaching to the test. Test scores stop being a good evaluation of a student's performance the moment you use them to evaluate the teacher.