2017 California school test scores: Why are they flatlining? • Hacker News Board

Re: 2017 California school test scores: Why are they flatlining?

#81

Post by dogruck » Tue, Oct 10, 2017, 4:13 AM UTC

In my eyes, standardized tests are bad because they will always shine a bright light on differences between groups.

We don't like to admit it these days, but humans are not identical. We can accept some basic differences, such as skin tone. We get extremely uncomfortable extending that notion to cognitive skills. To be clear, I'm not singling out any group -- I'm not linking skin tone to cognitive skill in any way.

So, we bring in standardized tests to level the playing field. At first, those tests are deeply flawed and filled with bias, making them especially harmful.

Then, we iterate, and remove those testing flaws. And we trick ourselves into thinking that will solve the problem.

But, suppose we arrive at the perfectly fair test. Suppose the test is highly predictive of, say, success at a 4 year college.

Now, what do you do if that test habitually ranks some group lower than other groups? To me, it feels like we are right back at square one.

Personally, I would prefer to focus our energy on helping each other out. If one person, or group, is falling behind, you find a way to help, and also to integrate each group into the whole.

I know I'm oversimplifying -- I just think standardized testing is a fruitless exercise.

Re: 2017 California school test scores: Why are they flatlining?

#82

Post by gizmo686 » Tue, Oct 10, 2017, 4:18 AM UTC

dogruck wrote: post #81

In my eyes, standardized tests are bad because they will always shine a bright light on differences between groups. We don't like to admit it these days, but humans are not identical. We can accept some basic differences, such as skin tone. We get extremely uncomfortable extending that notion to cognitive skills. To be clear, I'm not singling out any group -- I'm not linking skin tone to cognitive skill in any way. S…

> If one person, or group, is falling behind, you find a way to help.

And how do you know if a person or group is falling behind? To do this, you need some type of metric; and standardized testing is the best metric we have at scale.

Re: 2017 California school test scores: Why are they flatlining?

#83

Post by tristor » Tue, Oct 10, 2017, 4:21 AM UTC

dogruck wrote: post #81

In my eyes, standardized tests are bad because they will always shine a bright light on differences between groups. We don't like to admit it these days, but humans are not identical. We can accept some basic differences, such as skin tone. We get extremely uncomfortable extending that notion to cognitive skills. To be clear, I'm not singling out any group -- I'm not linking skin tone to cognitive skill in any way. S…

So, eliminate competition because it's unfair? That does not to me seem like a recipe for success on a global scale. If anything, the thing which has made the culture in the US uniquely attractive to so many people from all over the world is the fact that we revel in competition in a way most other societies do not. We love to compete, and are unabashed about keeping score.

Let's assume for a moment that we focus exclusively on helping everyone get ahead and eliminate standardized testing. What happens when people get out in the real world and get the shock they never had in school that they don't quite measure up? To prevent that would require a complete restructuring of society, it'd look wildly different. Among other things it wouldn't be possible with a capitalist economy by which the "best" products in the market are rewarded financially and competition is the basis of business.

Is that really the world you want to live in?

Re: 2017 California school test scores: Why are they flatlining?

#84

Post by home_boi » Tue, Oct 10, 2017, 4:22 AM UTC

labster wrote: post #15

MichaelGG wrote: post #4

Earlier quoted context omitted.

Is English the native language for most of California?

Yes. Most of the hispanics whose families arrived in California before the English speakers arrived have long since become native English speakers. As with most second-generation or later immigrants. Though I have heard tales of Californians being native speakers of Lisp, this is probably just a rumor.

Even if they are immigrants, knowing Spanish makes them half-way native English speakers because of the similarities between English and Spanish.

Re: 2017 California school test scores: Why are they flatlining?

#85

Post by home_boi » Tue, Oct 10, 2017, 4:27 AM UTC

gizmo686 wrote: post #63

ransom1538 wrote: post #36

Earlier quoted context omitted.

"This is the most important sentence in the article. It points to a problem in the test itself - California's students didn't just get magically stupider." Yes. Having gone through public education in California the tests were treated as a total joke. They had no weight to your grades and there was zero incentive to do well. Often you wouldn't even get the results. In elementary school it was common for a teacher to…

>They had no weight to your grades and there was zero incentive to do well. This doesn't seem like that bad of an idea. One of the bigger problems with tests is that they tend to become a terrible metric once people start preparing for them specifically (and test-prep can suck the oxegyn out of real education). If the students and teachers don't care about the test, it is much easier (but still difficult) to make a t…

> they tend to become a terrible metric

Not exactly. When a test becomes study-able, it becomes a good metric for how hard people try and how much time/effort people put into their studies/career. And that is arguably what people making tests were trying to or would want to measure.

Re: 2017 California school test scores: Why are they flatlining?

#86

Post by dogruck » Tue, Oct 10, 2017, 4:30 AM UTC

gizmo686 wrote: post #82

dogruck wrote: post #81

In my eyes, standardized tests are bad because they will always shine a bright light on differences between groups. We don't like to admit it these days, but humans are not identical. We can accept some basic differences, such as skin tone. We get extremely uncomfortable extending that notion to cognitive skills. To be clear, I'm not singling out any group -- I'm not linking skin tone to cognitive skill in any way. S…

> If one person, or group, is falling behind, you find a way to help. And how do you know if a person or group is falling behind? To do this, you need some type of metric; and standardized testing is the best metric we have at scale.

Ok, so you have your metric. What next?

They plug all of your stats into the function, and it says, "you will never be a good NBA player." With the NBA, we're comfortable saying "meh, not too many of those anyway -- moving on to high stakes poker -- never liked playing basketball anyway."

But when the metric is of something much more broad, such as "success at a 4 year college" we're not so comfortable doing that.

I just don't see what problem the standardized test actually solves.

Re: 2017 California school test scores: Why are they flatlining?

#87

Post by dogruck » Tue, Oct 10, 2017, 4:34 AM UTC

ProfessorLayton wrote: post #67

It’s not too surprising considering the rising inequality in CA. Wealth is the largest predictor of educational attainment. https://cepa.stanford.edu/sites/default/files/reardon%20whit...

Yes, but wealth is also correlated with aptitude.

I'm not saying our system is perfect. Plenty of rich bozos.

On the other hand, in a perfect system, to me, it seems that the highest achievers would tend to both: 1. Succeed in school and 2. Acquire wealth.

Re: 2017 California school test scores: Why are they flatlining?

#88

Post by gizmo686 » Tue, Oct 10, 2017, 4:39 AM UTC

dogruck wrote: post #86

gizmo686 wrote: post #82

Earlier quoted context omitted.

> If one person, or group, is falling behind, you find a way to help. And how do you know if a person or group is falling behind? To do this, you need some type of metric; and standardized testing is the best metric we have at scale.

Ok, so you have your metric. What next? They plug all of your stats into the function, and it says, "you will never be a good NBA player." With the NBA, we're comfortable saying "meh, not too many of those anyway -- moving on to high stakes poker -- never liked playing basketball anyway." But when the metric is of something much more broad, such as "success at a 4 year college" we're not so comfortable doing that. I…

We want a well educated population.

To do this, we have set up an educational system. However, education is a difficult problem, and it is not clear what works and what does not. What we need is some way of measuring how well things work. That is the problem that standardized testing is meant to solve. There are some tests that are intended to assist in accademic discrimination (eg. we only want good students here, so you need a good SAT score to get in), but that is not the usage we are talking about when we say California's are flatlining.

Google A/B tests moving a pixel on their home page. Compiler writers performance test optimization passes. Educators want objective feedback too.

Re: 2017 California school test scores: Why are they flatlining?

#89

Post by tptacek » Tue, Oct 10, 2017, 4:41 AM UTC

dogruck wrote: post #87

ProfessorLayton wrote: post #67

It’s not too surprising considering the rising inequality in CA. Wealth is the largest predictor of educational attainment. https://cepa.stanford.edu/sites/default/files/reardon%20whit...

Yes, but wealth is also correlated with aptitude. I'm not saying our system is perfect. Plenty of rich bozos. On the other hand, in a perfect system, to me, it seems that the highest achievers would tend to both: 1. Succeed in school and 2. Acquire wealth.

We're referring in this case to the wealth of parents, not of students. Students have very little control over which parents they're assigned.

Re: 2017 California school test scores: Why are they flatlining?

#90

Post by secabeen » Tue, Oct 10, 2017, 4:44 AM UTC

swivelmaster wrote: post #38

lr4444lr wrote: post #30

Earlier quoted context omitted.

So what's your recommended alternative metric? I don't doubt there is a good case to improve certain standardized tests, but that doesn't mean there isn't a valuable correlation even in a flawed one, nor that it's better to keep moving the goal posts into more subjective territory to explain away bad outcomes.

They're mentioned in the article. Graduation rates, college acceptance, stuff like that. Some of the countries with the best education systems in the world don't have these kinds of standardized tests. Also, nobody's suggesting moving the goal posts to explain away bad outcomes. If every school suddenly loses three points on average from one year to another on what is supposed to be the same test, the only common var…

Do you have a metric that doesn't require waiting until students are leaving high school to return results? Some of these scores are for primary school students. It's not very useful to their teachers to be told 5-10 years after they teach a student that the improvements in instruction they attempted had or failed to have an effect.