Gifted Issues Discussion homepage Forums Identification, Testing & Assessment Group ability test question

Hypothetical question here (not in re to my kids):

I know that we've discussed how many of us aren't terribly fond of the group ability tests (Raven, CogAT, OLSAT) as the ability criterion for GT program entry. I also know that I've read that the correlation with IQ isn't necessarily infallible in terms of a gifted child who may not test in the gifted range on a group test.

I am wondering, though, whether anyone has any opinion as to whether it is possible or likely for a child to test highly on part of a group ability test and not be gifted?

For instance, our local districts would id a child as gifted in a single subject (say math or reading) based upon a 95th percentile score on the math or reading portion of the CogAT. There are a lot of kids who are gifted per these guidelines. Probably 15-20% of kids score in the top 5% of some portion of some group ability test. Is this just a case of casting a broad net or is it possible for someone to be in the top 5% of quantitative ability on the CogAT (again, just a random example) and not be a gifted child? Would this same hypothetical child likely test in the top 5% of the quantitative portion of an IQ test then as well?

Possible, certainly. Indeed as we were discussing on some thread recently, it's entirely possible, even common, for someone to score in the top 5% even of an IQ test and then not score in the top 5% of a different IQ test. There is a strong tendency to think as though there is a "fact of the matter" concerning whether or not a given child is gifted - a notion of ur-giftedness, if you like - and there just isn't.

Likely? Correlation is the key here and it's a numerical notion, not one to which the term "infallible" is appropriate. I did recently hear it said that in the upper ranges there was no correlation between OLSAT and IQ tests; I think that must have been an exaggeration. Here is Wren here quoting a slightly more precise statement:
http://giftedissues.davidsongifted.org/BB/ubbthreads.php/topics/7395/OLSAT.html

It may seem odd that there could be a correlation between two tests overall but not at some particular level. As a thought experiment: it could be, say, that the very same group of children score above the mean [or, above the 75th percentile or whatever] on two tests, but that within that group of above-average children, knowing the child's score on one test tells you nothing about the child's likely score on the other (besides that it's above the mean, rsp the 75 percentile, which you already knew). This could be the case, for example, if the "basic" questions on the two tests were similar but the "hard" questions on the two were not.

Specifically concerning the tests you mention, I hope someone else knows. I had heard that Raven was well correlated with IQ, that OLSAT isn't in the upper ranges, and I know nothing about CogAT.

I guess that I ask b/c I am trying to understand what must be some misperception on my behalf. I'm trying to wrap my head around things that just aren't making sense to me.

I've said before that my dd11 tested in the 98th percentile (total) on IQ at 7. Her achievement scores have consistently been much higher than that and the tester did feel that it was somewhat of an underestimation due to her refusal to complete the block design subtest, which came out at the 25th percentile as compared to the other portions of PRI which were btwn the 99th to 99.9th. I believe pretty firmly that dd is a 99th percentile kid, but not a DYS level kid.

I am seeing kids who present to me as clearly not gifted who are ided as GT and who are scoring in the 95th percentile on at least part of the CogAT (that's the most common test used by both local districts). I realize that there are varying degrees of gifted and that dd is probably more than MG so she will, of course, look different than a MG child. However, I wouldn't expect the gap btwn her and MG kids to be so huge that they would appear average to me.

I am trying to figure out if I am just seriously misjudging some of these other kids who are in dd's classes or if maybe it just has something to do with personality on dd's behalf that she presents as more gifted than she is.

When I am saying that these other kids don't appear gifted to me, here's what I mean. I'll give a verbal example b/c there are some kids who have a LA GT id who've attempted to quiz dd what her scores were and proceeded to show dd their test scores so I have some hard data there. One of these children was in the 95th percentile on the CogAT verbal, and has a Lexile score that would be about 3 grades above grade level. That lexile is still 400+ points below where dd's was a year ago. The child speaks with grossly grammatically incorrect sentences regularly, just doesn't seem to "get" things that would indicate critical thinking capacity (I know that is subjective), and has a fairly poor vocabulary. Re vocab, for instance, this child regularly doesn't know what some of their spelling words mean and uses very simple vocab in speaking.

Reality is, it isn't an issue for me if schools are misidentifying, but if I am totally off and assuming that some of these kids are just being overidentified and they are actually gifted, then it does leave me wondering if I am also underestimating dd and I just don't think that I am underestimating her.

Would a difference of a few percentiles look that stark at the upper end? What must an average child look like, then? I don't think that I've met many average children if this is what MG looks like.

I am trained to give and often administer the K-BIT, which is one-on-one but 2 of the 3 subtests are multiple choice. It's just a screener, not a full scale IQ test, but it it sometimes used for gifted ID. (I use it as part of a dyslexia battery.)

Of course I don't know for certain which kids actually know the answers and which kids are just lucky with their guesses, but I have a general idea. The criterion to discontinue a subtest is when the kid has missed four questions in a row. Sometimes, a kid guesses right so often that when nearing his personal ceiling, he gets three wrong in a row and guesses the fourth one correctly which means I continue testing past his real ability. This can add a few points to the total score. The open-ended subtest is much more telling.

My DD took the OLSAT to qualify for services and just squeaked in with a 130. I dislike it as a qualifier because it was group administered, multiple choice, and timed. DD is the type that will conisder the nuances of two answer choices and lose track of time. I think a kid who makes quick decisions on what he truly knows and then also gets lucky on his guesses could overscore.

I did see something on Hoagies about the OLSAT having poor correlation with IQ for gifted kids here :

eta: I see that this same quote is in the thread you linked. Thanks . I also liked Dottie's comment on the other thread about gifted thinkers vs. high scorers on school ability tests. That is really the crux of what I am getting at. I am seeing kids who don't appear to be gifted thinkers to me who are apparently testing highly on tests like the CogAT.

That's aside from my local experiences b/c I don't think that test is used a lot locally. However, it does leave me wondering what one should make of a high score on the OLSAT then. Do we discount that b/c some highly gifted kids see negative correlations btwn IQ and OLSAT scores?

I have no help for you on the testing, but I can tell you that overestimating what "average" should look like is pretty common among parents of HG+ kids. It's part of that gifted denial.

I'd recommend that you at least consider the possibility that your DD is further into the tail of the Bell Curve than you realize. That may not be what's going on, but I think it's far more likely that you are underestimating your DD than it is that multiple kids are being overidentified as gifted. IMHO...

Maybe; I'm just not sure. I've had the GT coordinator of one of our schools tell me that she believes that we have a lot more "good students" than gifted kids in the GT programs, though, and we do, on avg, have btwn 15-30% of kids at most schools with a gifted id. I do think that is likely an overidentification issue. However, if 15-30% of local kids are scoring that highly on group ability tests, I am reconsidering that notion.

And, like I said, I don't think that more than a small percentage of the kids I've ever met seem noticably less able than these kids who are, apparently, 95th percentile kids. That just blows my mind b/c that would mean that avg is really, really slow appearing to me. One of the GT ided kids I can think of in particular strikes me as average, at best, and maybe somewhat below. I am shocked if this child really is a person with an IQ around 125 (I know that the CogAT isn't an IQ test, per se, but it is supposed to correlate). Of course, we are talking about one part of the CogAT, not the composite score and that might make a difference as well.

I think the group tests are crummy. I think they overID some kids and miss others. And we're also in a school system that IDs roughly 30% of kids as GT in one area or another. Some of that is probably accurate because of who lives here: corporate executives, professors, etc. Even so, I think that percentage is too high.

But I also know from personal experience that I tend to assume that MG kids are ND kids. We were in a toddler playgroup with several MG kids--who I now truly believe are MG!--and they seemed pretty "slow" to me back then in comparison to DS, who I thought was MG to HG.

I guess I'm suggesting that it may well be a combination of factors. But don't disregard your own potential to underestimate your DD. That's just personal anecdote talking, but I still think it is relevant and important.

It has taken me most of my teaching career (Latin, in diverse classrooms) to reorient to average. Average is frustratingly slow to me. Average is learning a skill in twenty-something repetitions (my GT population can sometimes master a skill in under five repetitions). My average kids consistently misuse homonyms and make obvious-to-me grammar and punctuation errors. Etc. My reasonably bright 125-ish kids don't all present as perfect (some are high achievers who do strive for details, though) but make fewer obvious errors and require fewer repetitions. More than five reps, though. My own learning speed is about 3-5 repetitions, so even my bright kids seem slow in comparision.

Sorry, this rambled.

No, that's not rambling. That's exactly what I am getting at. Dd11 has another very good friend who has a similar CogAT score (around 95th) to the other child from dd's school I referenced earlier. Dd's friend is clearly very bright. I don't question at all that she is a mildy gifted child. There is a difference btwn her and dd, but it isn't stark where one appears avg and the other gifted. One just appears more gifted than the other.

On the other hand, I am seeing these other kids with 95th percentile scores who do need 20+ repetitions to learn things and sometimes still don't get it, who say things like "she and me is going to play," when they are ided as gifted in language arts, whose vocabulary excludes words like "abdomen" or "superficial" (literally don't have any idea what they mean), etc. These are pre-teens, not young elementary kids. Is that really a 95th percentile child?