Also public schools in NY are zoned so you have to live the right place to get your kid into the right school.
That being said, I'm not sure either of these are a problem. The real problem (one of them at least) is that a school's performance is measured by the student's scores on standardized testing.
The selection bias of SA pales in comparison both purely by principle and in practice by the diversity of it's students.
And who care enough to move (showing a massive parental interest in their child's education).
This is why using test scores to compare schools, be it charter vs. public in the same district or public vs. public in another district, is inherently flawed.
With zoned schools you can have an interest in your childs education and have no way of getting in simply because you can't afford it.
The point is that those who claim that charter schools get to pick the good students and leave the bad ones with the public schools are simply wrong. It's the public schools in the rich areas that get to pick their students simply by whether they can afford to live where they do or not.
no, they aren't. Sorry, this isn't a matter of opinion, it's a matter of fact. Every study ever performed on this question indicates that you are the one who is wrong.
> It's the public schools in the rich areas that get to pick their students simply by whether they can afford to live where they do or not.
The existence of selection bias X doesn't preclude the existence of another selection bias Y... why would you think that's the case?
This is exactly what I meany by quasi-selection. The dat are absolutely clear on this question -- lotteries create a very nearly insurmountable selection bias.
> public schools in NY are zoned so you have to live the right place to get your kid into the right school
This is immaterial to the question of whether charter schools are more effective than public schools for reasons that are essential to their organization (rather than incidental).
Furthermore, I don't think this (or any) anecdata gets to the heart of my criticisms, which I'll explain further below.
The core of the claim you made above -- that a charter is one of the best performing schools -- is not even a reasonable claim in most states because Charters aren't subject to the same testing and reporting requirements as public schools!
In many areas, Charters are not required to report at all, or are allowed to self-report. Guess what? Public schools also do totally fanstically when they are allowed to sefl-report! Comparing self-reported data to data obtained through a standardized process is meaningless.
Again, I'm not saying this is necessarily the case in NY. Just that it is a real problem nationally.
No it's not immaterial whether someone can put their kids to a public school in Tribecca vs. Harlem.
Your comment started with "Again this is simply wrong", without specifying "this".
I assume you were refering to: "[charters are] selecting or quasi-selecting their student population".
So, my interpretation of your claim: "Charter schools that use lotteries out-perform, which means their performance is not attirbutable to student selection."
My response to this claim was that lotteries are a form of selection bias!, and that lotteries are exactly what I meant by "quasi-selection".
> Please argue against things I have said instead of things I never said.
I did respond to the substance of your comment. lotteries == selection bias, so a lottery school performing well isn't a sufficient answer to the problem of selection bias. (Also, if you don't want people to mis-interpret your comments, then don't open your comment with a dangling antecedent... saying "this is simply wrong" without saying what "this" is, is literally inviting people to interpret the thing to which you were referring.)
In other comments you've narrowed the scope of your claims to something about your child's specific school. If that's your intention, then Im not sure why you stared your comment with "this is simply wrong" when I clearly wasn't referring to your child's specific school, but rather was very clearly asserting a general statistical truth that applicable nationally and even internationally, regardless of the vagaries of a single individual event.
2. The back half of my comment was essentially: whether or not lottories are a form of selection bias, it's never-the-less the case that charter schools sometimes aren't evaluated in the same way as public schools, making it impossible to actually evaluate public vs. charter -- which is the central thesis of my comment.
> it's not immaterial whether someone can put their kids to a public school in Tribecca vs. Harlem.
Me: The charters vs public national comparisons are skewed comparisons in general because they're not being evaluated using equivalent metrics -- huge huge problem! Also, selection and quasi-selection bias.
You: Lotteries != selection bias. And there's another compounding variable -- income of student's parents!
Me: Lottories == selection bias. And yes, income or whatever is yet another thing we should control for when comparing public to charter. But in addition to controlling for these other variables, we need to -- at a minimum -- compare based on the same metrics. If charters don't have to take (the same) standarized tests, then controlling for every other variable in the world is pointless.
That you are right because you can claim that parents who are interested in their childs educations and apply to charter schools is the selection bias.
But this is not what this discussion is about. The claim being made was that charter school had an advantage because they got to pick the best kids and leave the rest for the public school system.
It doesn't and so the selection bias you keep insisting exist is really nothing compared to the one that exist in the public system where you are completely depending on how rich your parents are so they can live the right place and you are zoned for that.
Thats what parents do btw. Move to the good school districts. Those who can afford it. To any extent a lottery is selection bias it's simply not relevant or interesting for the premise of this discussion which is that the charter schools had an advantage.
They don't. Public schools in rich neighborhoods however do.
1. Just because X (here, income distribution) creates a bias that we have to correct for doesn't mean that the bias created by Y (here, self-selection for lottery participation) isn't also problematic form a statistical perspective.
2. There are a lot of reasons to believe that, in the case where we compare charters to publics, that the volunteer selection bias is more problematic than the income distribution bias in public schools. Namely, we pretty much already know how to statistically control for the latter, but it's a lot more difficult to statistically control for the former.
So, you're wrong. Period. Voluntary selection bias exists. This is just a scientific fact. You have to correct for it in statistical studies. Period. End of story. Do other biases exist? Yes. Are they more important? No, because that's a stupid question. Any sort of selection bias is a threat to validity, and the entire point of a bias is that it's hard to know exactly the effect it's having until you correct for that.
But, and I stress this because you seem to be ignoring it, the important portion of all of my posts has NOTHING to do with selection bias!
In many states we don't even have a meaningful statistic upon which to make a comparison between chaters and non-charters because charters aren't taking the same standardized tests. So even after perfectly correcting for bias, we still have zero hope of actually comparing charters to publics.
Fundamental take-away: any school taking taxpayer money should be required to take the same standardized tests as public schools. Otherwise, any and all comparisons between them are bunk.
I attended the best public schools in the county. Consistently listed in the top 10 nationally. My former public high school is effectively a private school. Just a few miles away are some of the worst performing schools in my state. Elsewhere, most schools are somewhere in the middle.
Similarly, some charters are above average, many are terrible, most are somewhere in between.
Overall, charter schools performance is no different than public schools.
All I want is for my kid to learn first of all reading, writing and math.
I don't even care about whether one school outperform another but there seem to be various studies showing different things.
My guess is that charter schools work a little better because of the Hawthorne Effect.
https://en.wikipedia.org/wiki/Hawthorne_effect
And studies show different things. According to this it's better in general. But to me thats less important you can probably find studies that shows other things too.
http://www.bloomberg.com/news/articles/2013-06-25/stanford-u...
Charter schools, in many states, occupy the absurd position of being able to claim that they out-perform local schools while selecting or quasi-selecting their student population and still not actually participating -- at all -- in the standardized testing that justified their existence in the first place.