Does a high GAI have the same rarity as the equivalent FSIQ would? On the GAI tables, the average is still 100 but, with less input to affect the score, wouldn't it mean it's easier to get a higher GAI?
My understanding is that it is normed so that the average is still 100 and the standard deviation is still 15. A GAI of 153 is in the middle of the 3rd standard deviation (145-160). A 146 on FSIQ or GAI is, theoretically, at the 1 in 1000 level. A 160 is, theoretically, at something like 1 in 10,000.
Because not many people score at these levels, there isn't a whole lot of data to back of the "theoretical" part, and I gather that pinning down the rarity of these scores is hard to do.
Also, keep in mind that the "map is not the territory". The scores are normed to reflect the rarity of the scores themselves. Just what it is that the scores are measuring is another matter.
Personally, I'm just beginning to come out of my own research/obsession with the meaning of my DS's scores. It seems to me that Dottie has a very sane take on it all :-)
Also, how do the extended norms fit in? Does the 145+ mean "beyond measure" with the GAI, even though there are extended norms that measure much higher?
I'm not sure what you mean. Did your son's report actually include a score cited as "145+". The percentile associated with a score over 145 always seems to be "99.9+", but I don't know what 145+ would mean.
If it's possible to score higher, then wouldn't it mean that the test is more sensitive at the upper tail and a score closer to the old ceiling of 160 is much less impressive since the test can go to 210, thus discerning between true 145s and those that were subject to ceiling issues?
I think that scores in the 140s are impressive regardless. Before the extended norms, however, you just knew it meant "off the chart". The extended norms give you an idea how far off the chart.
Finally a bonus question, lol- I read somewhere that scores between 17 and 19 represent ceiling scores. Was that only true before the extended norms? The tester specifically said that DS7 didn't ceiling any tests and his 18s were true 18s, not subject to the extended norms. How can a 17 or 18 be a ceiling if there were still more questions that could have been answered?
I suspect this means that more than 18 questions were asked, but that only 18 were answered correctly. I gather that after you get a certain number wrong, the tester stops asking. So if your DS got 18 of the first 20 questions right, and then missed 3 or 4 in a row, the tester would have stopped there after question 24. (But I'm quite fuzzy on this aspect of testing, so I could have this wrong)
If you DS had gotten 21 questions right, that score would still have been reported as an 18 for the purposes of computing FSIQ and GAI. If there were two tests on which he exceeded the ceiling, then the tester would computeand report an "Extended FSIQ" and "Extended GAI" in addition to the regular FSIQ and GAI that you got.
I hope something in there makes sense! Maybe I'm just over thinking this whole thing and it's much less confusing than I'm making it.
Thanks!
I think I know what you're going through. We're now 6 weeks post-testing and I think I've mostly gotten past my need to understand the numbers. Hope this helps!
