Gifted Issues Discussion homepage Forums Identification, Testing & Assessment I.Q. and grade levels ahead-strong correlation?

These results should not be taken at face value:

So how overblown were No. 1 Shanghai’s PISA results?
By Valerie Strauss
Washington Post
March 20, 2014

Furthermore, we know that Chinese-Americans outperform academically, especially in math. If the Chinese are beating us in math, to what extent is it due to differences in curriculum, teacher training, or innate ability?

... and/or family/cultural support/pressure?

I encourage you to read the actual report on PISA results for the US. It is not just China but a number of Asian and European countries who outperform the State's on average score but also having a larger number of high performing students and a smaller number of low performing students.

Just over one-quarter (26%) of 15-year-olds in the United States do not reach the PISA baseline Level 2 of mathematics proficiency, at which level students begin to demonstrate the skills that will enable them to participate effectively and productively in life. This percentage is higher than the OECD average of 23% and has remained unchanged since 2003. By contrast, in Hong Kong- China, Korea, Shanghai-China and Singapore, 10% of students or fewer are poor performers in mathematics.

There is a lot of interesting detail in the report related to how results are impacted by immigration (accounts for 4 % of the variance) and other factors including motivation and advantage.

All samples have flaws but this is still interesting reading.

Just an observation: I was checking out some of the top math kids in the US and there is a huge group from Bellevue WA. I checked out the schools in the area and the Asian population has spiked in the last few years, some schools are almost 50% Asian and the Math Programs in these schools always finish at the top in National and World Competitions.

Hmmm. I have issues with the WISC-IV. My son didn't score as gifted and the psychologist couldn't even calculate his IQ, yet he was ahead in a few areas as well. We haven't done a SB for him (I'm kind of disenchanted by the testing process for kids, to be honest. I wasn't tested until adulthood, and frankly I think there's more accuracy then). Also I've heard that the WISC is only good to about IQ 130 and anything higher is better served with the SB.

Enh. I think they each have their strengths and flaws. I do prefer the structure of the WISC-IV to the SBV, though, as it has a more coherent theoretical basis. From the sound of it, I'm guessing that your son may have had a highly skewed profile, which means that the FSIQ would not have been the best representation of his ability (probably the VCI or PRI would have been better), which would likely be why the psych didn't calculate an FSIQ. This does not mean that the instrument does not provide good information on giftedness; with a 2e or other low-incidence profile, there probably isn't an instrument out there that would really capture the full potential.

For HG+ individuals without 2e, testing in middle-to-late childhood is actually best (8-12 yo), as there is still enough range in the tests to be able to use extended norms. Scores may be more stable in adulthood, but ceiling effects become more problematic for the highest-scoring individuals. (Not an issue if we're looking at <145.)

No. When a six year old ceilings out on a subtest, the result puts them at 16 yo equivalent (from memory, paperwork is in sleeping child's room). And that's just not correct.

That's because the WISC age norms only go up to 16 yo. Therefore the highest possible age equivalent cannot top 16-11.

Obtaining the same score as does not equal functioning at the same level as.

And to the previous point: 10 yo children with IQ 130 scored at the same level as the median 13 yo under the old mental age/quotient IQ. (Note that this is not the same as saying that they are "as smart as" average 13 yos.) No contemporary IQ tests use the quotient IQ, nor have they since the 1972 re-norming of the SBLM. Although efforts were made to scale deviation IQs so they would have some resemblance to the numbers generated by quotient IQs, they really are not the same measurement system.

Perhaps what would make more sense (though it's not as readily accessible even when the data exists) is to look at Rasch scalings, which figure item difficulty in. (This is the basis of the W score and associated RPI on the WJIII.) E.g., those with an RPI in a certain range are expected to be receiving instruction in their ZPD. Those below need remediation, and those above need challenge. Unfortunately, the RPI doesn't spread the upper end of the curve at all well, since it was designed for spreading the lower end.

What a great idea from the perspective of understanding educational "fit" and appropriate level of instruction, though, aeh! If only it did work on the upper tail.