Statistical Errors Keep 4700 K-3rd Students From NYC 'Gifted' Programs
alostpacket writes "The New York times reports that statistical scoring by the standardized testing company Pearson incorrectly disqualified over 4700 students from a chance to enter gifted / advanced programs in New York City schools. Only students who score in the 90th percentile or above are eligible for these programs. Those in the 97th or above are eligible for 5 of the best programs. 'According to Pearson, three mistakes were made. Students' ages, which are used to calculate their percentile ranking against students of similar age, were recorded in years and months, but should also have counted days to be precise. Incorrect scoring tables were used. And the formula used to combine the two test parts into one percentile ranking contained an error.' No mention of enlisting the help of the gifted children was made in the Times article, but it also contained a now-corrected error. This submission likely also contains an erro"
All this "precision" to test against an arbitrary "90th" and "97th" percentile.
W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
They might as well file for bankruptcy. Their is no wrath like the wrath of a parent who thinks their child is gifted and talented.
I'm sure that a debate will emerge in the ensuing contents about the pros and cons of relying on standardized assessments as heavily as we do. From the summary, though, it seems as if the problem was not with the assessment, but rather the ancillary aspects of assessing. This doesn't excuse the mistakes, but it also isn't a compelling argument for abolishing standardized tests.
For what it's worth, Pearson is a for-profit educational publisher and assessment creator, but there are other assessment creators out there that are non-profit (e.g. ETS, the makers of the GRE). The entire assessment process is hard, and maintaining high-quality throughout is even harder.
Douglas Whitaker
Sad. It's amazing how much crap passes as science these days. We have people using the words of the field (six sigma, percentiles, normal distribution) and yet when asked about basic, basic metrics they completely reveal that they're just saying the words.
I wonder how people will say that stupid line, "x% of statistics are made up on the spot" or variations of it.
They want to have age in days?
Who defines age? They have to count from conception to be fair, since babies born before 8 months show delays academically if their early birth isn't accounted for.
Why are we measuring age with regards to giftedness? The age at which students are permitted to enroll in classes does not permit students born in September to enroll early, but if you are born in August you get to enroll, then when it comes to the standardized tests students who were permitted early entry get a lower bar with regards to entrance.
Since things are in flux to the point where a few days make a difference, wouldn't it make more sense to wait until there's some validity to the testing being done? As far as I know there's no validity to the notion that early testing leads to the right decisions being made. Some folks just develop early, but don't hit a particularly high mark, and some take longer to develop and ultimately to a higher level.
I remember when I was a kid getting screwed over because of my age, if a couple of months are that significant, then the testing shouldn't be done.
I personally think that Person is just a thinly veiled racketeering scheme. They should be disbanded. Their corporate leadership show highly immoral behavior, bordering on sociopathy.
Anyone else find some irony in the fact that the people deciding which kids qualify for advanced education programs couldn't get their math right?
My son was having trouble in his neighborhood school. The teachers/principal told us there was something wrong, likely ADHD or Aspergers... Broke our heart to be told this in a 5 minute 'parent teacher interview'... Anyway, after a psych-ed assessment, it turned out he just needed a gifted program possibly with some mild ADHD... The neighborhood school told us "Great! We'll just give him harder work. That'll keep him busy." but they already weren't dealing with his bullies, just brushing it off, and as we learned more about the gifted affliction, we understood that 'more and harder work' is not what he needed. He needed to be taught how his brain worked, how his brain was wired to learn in order to be successful as an adult... This is something that neither I nor his mother had when we were kids...
Anyway, the school board has a gifted program but they want _only_ gifted children and since his psych-ed report used the evil "ADHD" term, they rejected him... We found another school here (a Charter school, not private, still publicly funded, but more like an R&D sort of school) that catered to kids with multiple issues, including giftedness... Unfortunately, there were only 2 spots in his grade and more than 50 applicants... Coupled with a move to a new campus, an extra 25 spots opened up so my son got in. He's now in his second year there and this program has made a huge difference in his life... His first day at his new school, he came home and said "Mom! I've met my people!" ... He has a ton of friends at school, and is beginning to understand how his brain is wired... His teachers are giving him very successful coping strategies, and have imparted terrific insights to us about how to help him be successful... This has changed the trajectory of his life...
I can't imagine where we'd be were it not for this school... If we had been denied entrance, I dread think what state he'd be in...
I feel for the parents of these 4700 children, many of whom will not get the help they need...
Perhaps they should have used Excel and asked Carmen Reinhart and Kenneth Rogoff for some assistance.
So much for software not destroying peoples lives.
So these rounding errors matter because the statistical models would otherwise be 100% predictive? That's ridiculous. How does numerical age correlate with "giftedness" anyway? How does being a day older than someone else make you any smarter? People should be compared against their classmates/graduation cohort, not based on date of birth.
In the meantime, the truly gifted are hitting the library, doing their own thing, and pretty much don't need no stinking program.
In the pre-World War II era when Linus Pauling and Albert Einstein grew up, it was believed acceptably safe for a child to walk the streets unaccompanied. Nowadays kids are kept indoors over public hysteria over "stranger danger" and over poorly laid out, cul-de-sac-heavy street hierarchies that discourage getting from one place to another in anything but a passenger vehicle.
I find it hard to believe that a few days in the year change a child from gifted to ungifted. If it does, then these kids are on the extremely low end of gifted, and after a year they will even out with the rest of the kids their age. There was an /. article a while ago discussing how the gifted kids, that you see go to college at the age of 14 generally even out when they hit their twenties. Super geniuses are extremely rare. What normally happens is these kids are smarter than their peers, but not any smarter than your average adult. So they fly through highschool and college and end up at the same place everyone else does, just sooner.
"Cowardice in a race, as in an individual, is the unpardonable sin." --Teddy Roosevelt
My sister and I were both pronounced retarded by our elementary school for doing badly on the coursework.
We got sent to a special school where educational "decks" (cards) of lessons were available to us. We ripped through them because we were finally challenged and could work at our own pace.
That's when they figured out it was public school that was retarded, that we were just bored and public school pace couldn't hold our interest.
Then from (1976) age 12 to 18, after school each day I ran down to Florida State University campus and stayed there till midnight on their PLATO computers.
With limited resources, schools just cannot hope to teach at the pace of the fastest students without actually segregating them into faster-paced environments. Which pisses off all the parents of average paced students.
The world average IQ is 100, you retards! Haha!
One of the parents was a statistician. That's too funny for words.
What?
Would rather see students get access to faster-paced courses based on their percentile, if they're going to use that metric. Students in the upper-most values would have access to more seats, maybe a full course-load at 99%, and taper off the number of available seats to 1 somewhere around 87%.
Then, anyone "left out" at 86% is only really missing out on 1 faster-paced course than he would have taken already. It would be difficult to argue that missing 1 faster-paced course is a game-changer for his academic growth. Spend an extra hour studying every week to cover that gap.
we can make every student above average!
Look for a profit motive, right? Multiply the number of students by the profit they stood to make off of each student so classified. Someone close enough with enough knowledge of the processes and consequences of being misclassified to put these two things together has to do this.
Just the idea that you'd let a company like this supply AND measure students is a BIG mistake. If there's a chance to screw with the collection and/or processing of the data and one road takes you to greater profit and the other road takes you to lesser profit then you have a built in motivation to make "mistakes".
We're talking about a school text book company, a member in good standing of one of the most exploitative, manipulative RICO-ready industries in existence. One whose hands are filthy with the blood of the next bubble implosion - the trillion of so dollars of unserviceable student loan debt needed to cover, in measure, the massively inflated book prices that benefit not just the coke snorters at the top of this industry's corporate hierarchy but also the universities themselves.
The universities get a cut of cover price and believe me that total is a very, very BIG number. Thus the innumerable new 'versions" of textbooks which come out each year for the sole purpose of destroying the after-market for used text books. The company store was always a profitable idea but this one is astronomically profitable to all conspirators.
Its seems to me that the person(s) whom did the scoring where not from that advanced schooling..
in uni was statistics. The thing about statistics is its often just common sense. 1 mistake = careless, 2 = incompetent, 3 = corrupt
Unless it is exceptionally different in NYC, it's not a huge loss. When I was in the TAG program on the west coast, I found that all it really meant was the material you studied was *slightly* less idiotic than the mainstream - but still fairly underwhelming. For example, in the English class we were reading Hemingway, while the mainstream reading class were reading Jurassic Park (yes, that highly regarded classical masterpiece of classical literature about a park of dinosaurs). In the long run - like most things in high school, including your actual "high school record", nobody beyond high school actually gives the slightest fuck about it. At no point in life has it ever been relevant to anything or asked about or. Well, until this Slashdot post. So . . yeah, I guess it's good for that. Hurrah.
Also, that they let me into the program sort of proves how irrelevant an stupid it is -- even when they aren't fudging statistics. :P
END government at all levels having ANY involvement in personal educatlon through the age of consent. There ya' go. PROBLEM SOLVED!
The term "personal education" is used deliberately.
Plenty of crapitlist money floating around out there which new Carnegie corporations trying to "buy their way into heaven" (MicroSerf) will spend "for the children".
Single, no children so my "dog in the fight" is MY MONEY, stolen by a corrupt bureaucracy for an even more corrupted and just plain government by thuggery (see proxy voting prior to 1995 in the US House of Representatives). My siblings children are SCREWED!
My own experiences were that I probably would have been diagnosed as attention deficit disorder at a minimum, and border Asburger if such crap had been used in the 1960s.
They tested X children. The 10% that performed best were awarded with the stamp "gifted". The 3% that performed best were awarded with the stamp "especially gifted". If they now give 4700 children more the stamp "gifted", they will have more than 10% of the children with a stamp stating they belong to the top 10%. That can't be right, so the current results aren't wrong. I'm sure every school wants all of their children to perform "better than average", but in reality, if they all perform better, the average goes up and you still end up with half of them performing worse than average. The same applies here. The top 10% is the top 10%, no matter how you rate.
I was promised a flying car. Where is my flying car?
Statistics is the art of lying to yourself that you understand what should happen! The problem with statistics is that in general you just ignore mass amounts of data to try and prove a point, it would be like measuring a system and throwing out any strange readings. There is a much better system already in place that deals with the problem in statistics and that is to just measure or record the results of a finite population and only release results about that and only that population. You'll have the EXACT numbers and you wont be guessing or making assumptions!
"the city has been unable to control the explosive growth in high test scores, which coincided with the growth in test preparation services"
"There are three kinds of lies: lies, damned lies, and statistics." Attributed to Benjamin Disraeli by Mark Twain
standardized testing can passover people who are smart but are not good at taking tests.
My uni uses Pearson software (MasteringPhysics in this case) and I once forgot my password. I used to restore password link, and lo and behold: they send me my password in plain-text. If they're incompetent enough to NOT hash passwords I do not think they have any right handling anything IT related.
When I as a kid I got kicked out of the Gifted programs for lack of funding. There were only 9 of us that bothered and that wasn't deemed enough to pay for our teacher.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
Go away AC.
Do your own homework and write a 500 word essay on how educational theory has changed in 100 years. Pick any two of the hundred factors that have changed.
I tested in to the high 99% in K-12.
I spent more time in detention for "checking-out" during the rote learning phase of the Prussian education system, than I did in "TAG" or "FOCUS" or whatever Will Hunting environment they were trying to create for me at the time. These gifted programs are just petting zoos for educators to jizz their pants over the fantasy that that they might inspire curiosity in the next Turing.
http://en.wikipedia.org/wiki/Prussian_education_system
They withheld these programs from me as if they were some sort of scooby-snack they could reward me with for shitting out golden eggs. They were butthurt that I wasn't slaving away for the benefit of their metrics and was a "bad apple" that was spoiling their ability to farm money from idiots.
You get rid of merit based scholarships and why should anyone care about their GPA before High School? Or even in High School for that matter? Maybe graduate school is better, but I doubt it. College so far is an extremely disappointing high school 2.0. At least I've learned how to make the best of it.
TLDR: If you're not 98%, you should have tried harder. J/K, effort has nothing to do with it! Consider yourself lucky that you don't have the magnifying glass of high expectations turning every interaction with educators in to a lesson in the look of disappointment &/or the feeling of failure.
Do these passover people attend synagogue?
The whole of my early education can be summed up by "too easy, I'm bored". Cool, right? Well, no. Because I never had to study for anything, I didn't learn how to study. So when I eventually got into university I failed for the first time. I couldn't handle that, because I never learned how to handle that. So I flunked. Not because I'm not smart enough, but because I never had been challenged up to that point.
So, if I got a say in this, I would challenge *all* the kids, from an early age. Make sure everyone gets to the top of their capabilities. Some will top out at a lower level than others, that's OK. Give all the chance to grow and benefit of the educational system.
Don't try to fit all kids into the same mold. It doesn't work. Not everyone is the same, it is high time the educational bigwigs acknowledge that.
I'll forward an e-mail that I received recently, which should fix any percentile dysfunction problems.
I'm not surprised that there was an "error" that denied a large number of students the label gifted. In the New World Order, there won't be gifted people, just people who fit society and those who don't. Those who don't will be eliminated for the good of society. So, the gifted, if they are truly so, will quickly learn to act like the rest of the schlubs that will make up the majority of society.
One can only hope that there will be enough of them around to help when the new world order collapses on itself.
The real silliness is that educational opportunities are assigned by some central authority. We need much more of a free market in education. Give vouchers to people who can't afford it, but let parents make choices for their kids and let schools specialize in lots of different kinds of kids and programs as parents demand.
Agreed. (in part)
Passing someone onto a course level, which they cannot be realistically expected to succeed in, repeatedly, is more akin to punishment in my mind.
It is probably due to the structure of education, where there is a period of one year, that leads to the view that being held back is punishing. If early education were more modular specific skills could be targeted much more efficiently. This might require a larger scale to be cost effective, although it would also permit greater specialization on the part of instructors.
That being said, the grandparent does have a point about the allocation of resources. The public school system needs to serve the basic educational needs of the majority of the USA. This is critical for a rational effective democracy. It is also critical in a global economy where the value unskilled labor is rapidly sinking below a level where people can sustain their economic independence in a "developed economy" country.
The USA is already capable of attracting the "best and brightest" people from all over the world due to economic opportunity and high standard of living. It seems unlikely that the public school system will be able to effectively serve the educational needs of a geographically dispersed minority population "gifted students". I have already seen quite a few comments from slashdot commentators that seem to support the claim that "historical gifted student programs fail to provide adequate educational opportunity to highly gifted students. It might be better for gifted students to graduate quickly from the public school system, and then purchase higher education in the private market, perhaps with the aid of some tax incentives and need-qualified public scholarship programs.
This may eventually lead to a kind of scenario where the public school system is trying to "trap" students as long as possible in order to generate revenue, but perhaps with proper incentives this risk could be minimized.
It can also reward people that are just really good at picking the best answer out of 4.
On state tests in middle school and high school I was always in the 99th percentile, I was also failing all of the advanced classes the TAG program pushed me into.
If you gave me a test with multiple choice answers I'd probably do well. If you gave me the same test, but make me answer it myself (instead of making me sniff out the wrong answers) I'd probably score astronomically lower.
Of course part of my problem was that I had horrible study skills, hated homework, and just didn't want to be there.
I don't think it is okay to already start sorting out which students are gifted are not at elementary school age. For example, my son is now a student at Georgia Tech, but he was never admitted into the gifted program because they said he scored too low in the verbal reasoning and creativity sections. He was stuck in the slow classes and lost interest in learning until I took notice and began teaching him myself. I believe that all students should simply be given a choice to taken the "gifted track" and see how they do.
Shout out to everyone who's less gifted...
Joel Spolsky nailed it years ago. Whenever you change your arbitrary quality or quantiry metric, people change how to game the metrics by which they are rewarded.
That AC is more right than you know, because the United States (and many other countries) is still mired in the 19th century's broken and degrading Prussian "factory for minds" system. Practically all of your "advancements" in "educational theory" amounts to shuffling the deck chairs on the Titanic, as the central tenet of the Prussian system's Industrial Era-philosophy is that minds can be "standardized" like machine parts coming off an assembly line is ultimately rotten to the core.
You solve the street hierarchy problem for walkers by placing footpaths between the ends of opposing cul-de-sacs.
That's how it should be done for new construction. But in practice, a city would have to retrofit existing cul-de-sacs to include such paved trails, and with the "taxed enough already" mentality that affects my country, I'm not sure where a city should find the money for that.