That makes sense, incog. Yes, on her WISC IV test result it does say full score 99.6% (verbal is 99.9%). Crazy...fractions of a percent make a big difference!

Sounds good Incog! Minor point, and it seems so obvious to me at THIS time, but I'm pointing it out because I truly didn't get it way back when. The percentiles are against the norming sample only (vs "all kids who took the test"). This is significant, because it lowers the test population considerably. There are very generous norming samples, but even sample sizes in the thousands don't allow for many kids in the top percentiles, kwim?

Processing issues aside, the ceilings were "lower" on the WISC-III than the IV. By this I mean that it was easier to get 19's on the III than the IV. I have quasi-proof of this somewhere, but don't ask me to prove it. Professionals taught to accept The Almight Test Norms won't necessarily buy that anyway.

Oh, and the study about the 123 scores was Harcourt's very own "validation report". This is a standard I think for IQ test revisions. It's not part of the norming process, but done afterward on various subgroups, GT included. Scores on the SB-5 were similar (in the 123/124 range for "GT" kids). It's never very clear though how they identified those GT kids for the study.

That is an interesting tidbit there Delbows!

Disclaimer (I love disclaimers!): Most of my thoughts on this entire thread come about after getting "lower than expected" scores on the WISC-IV, and having no first-hand data on the SB-5 (despite all the articles I've read on both tests). In fact, DD13 and DS8 have the same GAI on the WISC-IV. While I certainly love each child equally (another disclaimer for those lurkers who think I favor my son), they are NOT the same LOG!

So...take my thoughts with a pillar of salt. I think GT kids can get unreliable test data in both directions in that upper tail, and that to truly figure out LOG, other factors need to be considered.

DS's testing experience on the WISC-IV though was far from ideal, and includes comments suggesting it's an under-estimate. If anything, DD13's is possibly an over-estimate.

I thought that was interesting too. It may be similiar to having a "preview" of the EXPLORE test. I know a group (school) that he said was included as part of the "validation report" (thanks for the clarification)are DYS level. They have to have tests at or above 145 and strong recommendation, ect. to attend the program.

A third measure, besides score and percentile, is the age equivalent. The author of the above study brought up the perceived discrepancy between standard score/ percentile and the age equivalent. Again, this rings true in my son’s case. Even though he just cleared the traditional gifted cut-off, his age was equivalent was 55. He tested the same week that he turned 10. I'm not too sure about this number either.

Given all I've forgotten by 42, 55 might not be that great of a score, !

We never got AE's with any of the IQ tests my kids have taken.

Thanks Dottie and delbows for clarifying. And thanks for being kind in responding to my first attempt at " I think I know what I'm talking about"!

If I remember correctly concerning the Harcourt Study(thanks, I'm great at pulling data out of the air and not remembering where I got it from), the kids chosen had previously scored 130+ on the previous WISC.
Does anyone have the link, I could probably find it if needed.

Is that what you meant by not sure how the children were chosen?
Also, Dottie, I really am pretty new to this and it was explained to me "all the children who took the test".
Could you explain the term norming sample. Remember, I'm not the math/statistics person and someone told me there are no stupid questions..........right????
Please don't say only stupid people

Thanks,
Incog

True. The 99.9 percentile for AE may be 26 or 27.

"Norming Samples" are typically made from an appropriate sampling of the national demographics.

For the truly bored, here are all the WISC-IV Technical reports.

http://harcourtassessment.com/HAIWEB/Cul...p;Mode=resource

Lots of interesting info in all of those. #3 talks about the validity tests, but I'm not sure it's clear about how those "GT" kids were selected.

I'm not sure if the norming sample info is in any of those, but I also found this...



All of the scores, percentiles, etc are based on this sampling of children (I thought it was >3000 though?)

FWIW, the SB-5 has it's own barrage of reports!

I've thought a lot about this study since I have a kid who has been tested on the SB5 and has a score that is PG by the Ruf/Hoagies scales but is below a 145. I really wanted this study to prove that he "should" have been over a 145 *if* we had the "right" test!

But I don't think it does. There are a lot of factors that go into this score difference.
1) The SB5 was designed to give a score that is about 3-4 points lower than previous tests to compensate for the Flynn effect (the observation that IQ test scores increase over time). I think one of the things that is contraversial is that this effect has not been shown to exist at the tails of the curve.
2)The tests are measuring different kinds of intelligence, different skill, knowledge, processing etc, so a child with the skills to do well on one test might lack the skills to do well on the other. This does not mean that one is more or less smart, just differently smart.
3) Regression to the mean. The statistical principle whereby outlyers on a test tend to get scores closer to the mean when the test is repeated and
4) the kids got older between test 1 and test 2. They may have hit fewer test ceilings on the first test and more ceilings on the second because as they got older they were working at the higher levels of the test.

What we need to really prove the difference (to go along with this study) is the opposite study--the one that tests kids first on the SB5 and then retests the highest scorers on the WISC3. You might confirm that the SB5 consistently gives lower scores. But one might find that the kids that do well on the SB5 don't do as well on the WISC. You just don't know until that study has been done.

I do know that when I discussed all my ideas about why DS's score showed he was PG with the DITD counselor after we had not been accepted, she just said, "well, we have plenty of kids who do score over 145 on the SB5." For those of you applying, though, with an IQ score below the cutoff, keep in mind that we had invalid AT scores (we had thought they were accepted subtests but they weren't). For kids with high enough AT scores, they might look more closely at those close, but not quite, IQ scores.

There ARE a lot of studies being done. What I want to know, is how I can get my kid signed up for them (and the free IQ tests that go along with that!!!!)

Your post makes sense ACS. Another study I'd appreciate is one that compares the WISC-IV to the SB-5, not just the III.

I really wish DS had SB-5 numbers (and WISC-III numbers, ), so I could contribute more personally.

(One of these days I'm going to give up on the Ultimate IQ Dream, and go with the achievement we are consistently getting.)