My school district is changing its gifted identification tests. It has been using the Raven for about the last 20 years. Two tests it is considering are the Cogat and the NNAT. (Also the WISC, but I can't see that being a realistic contender because of the cost. I am assuming it would cost a lot more because it's individually administered. Am I correct?) I was just wondering what would be some of the pros and cons for these two tests. I have the opportunity to give my input and I'd like to know what I'm talking about.

CogAT:

-measures aptitude in more domains, which provides more options for selection criteria (C/V/NV/Q).

NNAT2:

-believed to be more culturally-fair (though some research disputes this).
-usually shorter

Thanks, aeh. I like the idea of having more information about the different domains of aptitude. I think that can be useful for every child that takes this test, not just to identify for the GATE program.

My son's district uses both the NNAT and the cogat, so that's also an option. They screen everyone in second grade with the NNAT and those who do well (they select the top 5% in the grade - in my child's school this includes scores below the top 5% nationally) are nominated for GT evaluation, which includes cogat testing. Parents/teachers can also nominate for evaluation regardless of NNAT scores. Kids can qualify with high enough NNAT or cogat scores, plus other criteria. They also identify kids as gifted in math/science (high quantitative cogat score), Language arts/social studies (high verbal score), or both (high overall). At this point they don't really do anything with that information though as it's a general enrichment program and not subject specific.

They used to just use cogat testing for the entire second grade (and other grades as nominated), but I'm told that the testing took too long so they switched to screening with the NNAT first.

FWIW, I have yet to meet a gifted kid IRL for whom the CogAT was accurate. I know a few kids who did not score highly enough on the CogAT to qualify for a gifted program (in our district, 98th percentile), but private one-on-one IQ testing showed otherwise.

Accordingly, I think it's important to recognize that not testing highly on the CogAT does not mean a kid isn't gifted, and if CogAT is the only method of initial identification, a significant number of kids may be missed. I assume my view may be particularly applicable to students with twice-exceptionalities (using that term loosely), especially relative processing speed issues.

Snowgirl, I've actually experienced the opposite. The kids that I know in our gifted program who are there despite low CogAT scores but because of high private test scores have had lower scores on achievement testing such as MAP in later grades. For my kids, the only weird thing about the CogAT was that they scored slightly lower on the verbal section than on the non-verbal section, but later IQ testing showed verbal to be the highest.

I don't know the current MAP scores for the other kids I had in mind (none of whom participated in the full-time gifted program in our district), but as for my own kids, on the CogAT in 3rd grade they scored lower on the quantitative and nonverbal sections than they did on the verbal, which also is the opposite of where their strengths lie - that's funny that you also saw a reversal.

In any case, their CogAT nonverbal and quantitative percentiles weren't even anywhere remotely near the 90s (one was so low that the teacher laughed at the inaccuracy). So far, they are still scoring well inside the 99th percentile on the math MAP at the end of 6th. (They did not participate in the full-time gifted program either; they've been in a charter all along)

Was this using a timed version?

Yes. Timing was my obvious concern before the test, as both kids have low coding subtest scores (single-digit percentiles). It's been a few years obviously, but if I recall correctly, they both said that they finished it (I'm not sure whether that's true). So, I'm not sure what the problem was.

Gifted does not equal high test scores particularly if there is something else in play.

Despite the limitations of these tests, I'm not aware of any other group tests that are any better at gifted identification. Some schools also use achievement tests, typically the MAP tests I think, as part of their identification process.

I like this model. Do think this works well?

I think it works reasonably well, though maybe they should be testing the top 10% on the NNAT with the cogat to ensure they don't miss many kids who would potentially score as gifted on the cogat. The counselor told me that they were thinking of changing the universal screening grade to third grade, which makes sense to me as I'm told these tests are more accurate in older children.

Kids can also test for GT as early as kindergarten by parent or teacher request. I think they are only given the cogat in those cases, not the NNAT. They also can only be evaluated for GT twice in elementary school. It looks like roughly half of kids in my son's school are identified before the second grade screen, maybe 40% more in the second grade and just a few more upper elementary.

I think one of the benefits of not administering the cogat to everyone is that you don't have an entire class of kids taking the test at the same time in the computer lab. When my son took it in the second grade, it was him and 3 other students. That has to be a lot less distracting.

In our district they currently do universal testing in second grade with programs starting in 3rd. They are thinking about changing testing to 3rd, but us parent advocacy group leaders want to keep it at second. We already feel like that's late because kids can loose their spark even by then. My DD was utterly deflated in second grade.

yeah, I think it's a balance in reaching kids early enough and accuracy in identification. My son's scores went up 30 points from first grade to third grade testing. Obviously I can't prove which is more accurate, but given his abilities, I would say the third grade test is closer. I recognize that his huge leap in scores is not typical, but smaller, yet significant changes probably are common. Some kids are going to do well at any age, but I think maturity makes a difference for some kids. I don't know what to say about kids who initially test gifted and then go the other direction when tested later, but I'm sure that happens too. The format of the cogat test changes in third grade so that causes some of the differences.

We also have a two tier gifted program, "regular" GT is cluster grouping with differentiation (that we never saw any of other than spelling words) and 1.5 hours/week of pullout enrichment. There is also a highly gifted program that is a self contained magnet class for the district with acceleration and added depth of instruction. Testing for the highly gifted program is only done on current GT kids in the third grade (program starts in fourth).

Yeah, some kids apparently have had trouble filling in the bubbles using the current test. Those who score close to the cut off can retest in 5th and they usually do better that time around. This problem may be alleviated, though, if switching to a computer based test.

The same test is used for identification for our HG program which unfortunately has become more of a high achiever program. I'm working on fixing that too,but that's for a different thread.

Our school does the computerized test so it's not just about bubbles. The child has to be mature enough to concentrate on a test for as long as it takes to complete. Older children typically have longer attention spans. Also, from what I understand, the proctor administers each question and then waits for each child who is testing to select their answer to move on to the next one. I can see how this would be distracting for some kids, especially for those who answer quickly.

From third grade on, they work through the test on their own within the specified time limit.

At any rate, I think it's good to have the option to take tests both in early and later elementary, but obviously not so much that kids are testing every year. My child took the test in first grade (parent/teacher referral - average scores, not admitted), second grade (passed the NNAT screen and nominated for additional testing - was admitted due to high quantitative score), and third grade (all GT students screened for HG program). For some reason they test all kids admitted into the HG program again in fifth grade, so he will have taken the cogat test 4 times in elementary school. At least he gets next year off.

This article is relevant, but note that the Lohman authored the CogAT.

Comparing CogAT, NNAT, and the Raven
by David F. Lohman
2008

I did read that. Interesting, but taken with a grain of salt.

That's an interesting point. With any group test I suppose there would be some kind of distraction or pressure to finish when everyone else does. Individual testing would be best, but I don't see that as a possibility in my school district. We have suggested, though, that there be additional testing for those that ceiling any screening test.

I think the interesting thing in regards to that article and your situation of changing to the cogat and/or nnat is the possibility that a significantly smaller portion of kids are going to be identified as gifted. I believe aeh has also said that the ravens norms are outdated and would now over identify giftedness.

Also note that there are new versions/norms now for the NNAT (NNAT2) and the Cogat (form 7 - I think that article was based on form 6). I have no idea if Ravens has been updated. Changing tests no doubt changes some of the kids who will be identified/not identified. It's weird to think about, but I guess it's unavoidable to a certain extent.

The CogAt and NNAT norms have been updated, with noticeable improvement for the NNAT2, but the Ravens' have not.

There have been a number of attempts to clean up the norms for the Ravens over the past few decades, but most of them consist of convenience samples. In the USA, there are 1984/1986 children's norms derived from 22,000 samples collected in ten regional sites (aka, convenience samples), with no true national norms, just some extra-smoothed approximations. There are also 1993 adult norms from---Des Moines, Iowa?! Pearson waves its hands and claims that Des Moines is pretty representative of the whole USA, right?

So the children's norms are 30 years old, and weren't population representative to begin with, and the adult norms are nearly as old, and apparently representative of a single midwest city.

I oddly think that having less GATE identified kids in the school district may actually be more beneficial to getting the services they need. This school year we have 27.7% of the student population being identified for GATE placement (98%ile and above). They say we are a "smart city" but that seems ridiculously high. 5.4% of that number is identified for the highly gifted program (99.9%ile). Perhaps having numbers that high actually devalues their special needs.

HID--we have the same issue in our big city. It actually has made me wonder if the tests are normed on a very low achieving group. Our city allows kids to be tested yearly and also allows for appeals using private testing, so I've also wondered if the odds are just stacked in favor of kids who are tested multiple times.

SES does play a big role in the number of gifted kids, but 27.7% still seems high. Our district is pretty average for Texas, and likely a bit below average for country as Texas is not a high achieving state. Our district identifies about 10% of kids as gifted eventually, and number of kids id'd varies quite a bit by school in the district depending mainly on how wealthy the area is. The higher SES schools identify ~16% of kids. I know this from looking at our state testing data, one of the groups they stratify by is gifted and talented. I've also noticed that the gifted kids at the higher SES schools tend to score higher on average than gifted kids at lower SES schools on the state tests too. Very interesting.

They use a cut score of 130 on any section of the cogat or nnat, but kids in the 120's can be accepted too if they do well in other parts of the evaluation. The form also states that they take into account the educational background that the child comes from, which to me means SES. So really not very stringent standards and still only 10% are accepted overall.