Re: more errors by Pearson; is it really possible?
- Breaking: a parent wrote the following to our neighborhood listserv:On a serious note though...what's really upsetting is that we were one of the 300 this time too and now the new (3rd) score result form is riddled with errors and patently incorrect numbers. Here goes another hour long call to the doe...
- No, that's easily possible (not so unlikely). Percentiles from two sub-sections are not multiplied. You could even be in the 99th percentile overall if you are in the 90th percentile for each of the two sub-sections.
Say you start with 100 students. Student "Sandy" scores top 10 on both parts (90th percentile), but 49 of the remaining students do well on only part A and poorly on Part B, and the other 50 do well on Part B and poorly on Part A. Their overall scores are all middling, even for those 9 who score above Sandy in each sub-section (total 18)
Sandy's overall score is #1 and that puts Sandy in the 99th percentile.
Sorry for the math lesson.
--- In email@example.com, Rachel Paster <rpaster@...> wrote:
> something else that defies the definition of percentile is a kid scoring at
> the 90th percentile in both of the two sub-sections of the test, with a
> final score of 92nd percentile. but I'm not a mathematician, so maybe it's
> only highly improbable.
> On Mon, May 13, 2013 at 3:43 PM, ccssimath <ccssimath@...> wrote:
> > **
> > 10% of all kids scoring in the 97th percentile or above is mathematically
> > impossible, or more simply, it defies the definition of percentile. It's
> > akin to Lake Woebegone's children all being above average.
- I can't imagine any way any test can measure soething called "potential". First of all, potential for what? Good, evil? Making a difference for others? Making a lot of money? And if you decide which of these--what would you know at age 3 or 4 that could predict such???On May 13, 2013, at 9:31 AM, Edith Baltazar wrote:Carrie, thank you for sharing the articles. We all have to be questioning these G&T tests--such high stakes,yet it's treated as a parlor game--a kid could test really high and get no placement anywhere.So there is a huge problem with the content of these tests.I say parents deserve a G&T after the debacle with the G&Ts.Here is an opinion piece from today by the Daily News, no less!http://www.nydailynews.com/opinion/final-exam-article-1.1340975On Mon, May 13, 2013 at 9:12 AM, carrie mclaren <brooklynite282@...> wrote:On Mon, May 13, 2013 at 8:52 AM, Leonie Haimson <leonie@...> wrote:Actually in certain communities in the wealthier parts of Manhattan, 50% of the students tested “gifted” – to DOE, meaning in the top 90th percentile or more. The whole concept is absurd – esp. as studies show that most of kids who test “gifted” at 4 no longer test “gifted’ a few years later -- despite all the advantages of being in a gifted class.
For more on this, see:
The New Child-Testing Craze
Alternatives to IQ tests are suddenly all the rage. But they’re even worse at predicting kids’ futures.
(This piece STARTS with the premise that IQ tests for 4 year olds is demonstrably absurd)
The Junior Meritocracy
Why kindergarten-admission tests are worthless, at best
(whose spazzy 4 yo just bombed the gifted and talented test)