Gifted or test prepped?

Test prep for four-year-olds keeps escalating in Manhattan, reports the New York Times. It’s a game played by well-to-do parents eager to get their kids into public gifted programs or into selective private schools.

The New York City Education Department changed part of its admissions exam for its gifted and talented programs last year, in part to minimize the benefits of test prep. A test prep company “posted the news with links to guides and practice tests for the new assessment,” reports the Times.

The day Pearson announced changes in the exam used by many private schools, another company explained the changes in its blog:  “word reasoning and picture comprehension were out, bug search and animal coding were in.”

Schools worry that intensive test prep has made the admissions test invalid.

Nearly 5,000 children qualified for gifted slots in the city’s public kindergartens this year, double the number five years ago.

Natalie Viderman, 4, spent an hour and a half each week for six months at Bright Kids NYC, a tutoring company, working on skills like spatial visualization and serial reasoning, which are part of the Naglieri Nonverbal Ability Test, or NNAT 2, the new gifted and talented test. She and her mother, Victoria Preys, also worked every night on general learning, test prep and workbooks, some provided by Bright Kids.

Natalie’s brother, a Bright Kids graduate, tested into a gifted program. Natalie just missed the cut-off for a gifted school that uses an IQ test but hasn’t heard if she’s qualified for a gifted program that uses the NNAT 2.

About Joanne


  1. Sounds like more ‘teaching to the test’ rather than actual gifted student ability (IMO, we use the Gifted term way too frequently in schools) and I’d prefer that we restrict use of the term to perhaps the top 5% of all students in student body.

    I mean, look at the issues with valedictorians, how can a class of 800 students have 30 valedictorians, after a while it kind of demeans the honor (which is why a lot of high schools and districts have eliminated the position at graduation) and instead recognized the top 1, 5 and 10% of the graduating class.


    • Public schools are creating the problem because they refuse to offer homogeneous grouping and anything that would challenge the kids at the top. Even if they’re not specifically “gifted”, bright and motivated kids deserve to have their needs met, too.

    • palisadesk says:

      “I’d prefer that we restrict use of the term to perhaps the top 5% of all students in student body”

      Don’t forget that exact criteria vary considerably. My district restricts “gifted’ classification to those in the top 1% of measured cognitive ability. On the WISC-IV they need a GAI (General Ability Index) of 99 th %ile or higher, and a similar percentile if a different measure, such as Ravens, Stanford-Binet or WJ-III, is used.

      Eliminating some of the subtests from consideration does make it more difficult to “teach to the test.” While vocabulary and knowledge subtests can be coached to a degree, the actual reasoning tasks are less amenable to this. It’s likely though that some test prep does put individuals in the boundary areas over the cutoff point. For example, the tasks on the Ravens are not language-based, but they are usually verbally mediated, and children (or adults) can be taught strategies to think them through. The performance of the individual is still constrained by his/her actual reasoning ability, however.

      So “coaching” might get a 97th percentile individual to score at the 99th, but not a 60th percentile individual.

      If people actually get access to the test *items,* however, this can skew results significantly. There have been documented instances of children been taught to score high on standardized IQ measures by memorizing the correct responses. That’s why security on these tests has been beefed up.

      • Vocabulary is, hands down, the least amenable to coaching. If it weren’t, SAT verbal scores would be much, much higher, and the old GRE wouldn’t have had just 2% scoring over 700 on the verbal.

        • palisadesk says:

          We are not talking about secondary students or GRE takers, we’re talking of 4-7 year olds, approximately. Vocabulary absolutely *can* be taught (well above the student’s cognitive level) to young children. They are not, even in most cognitive testing, required to do advanced things like analogies with this vocabulary. Also, the amount of vocabulary expected of a 4-year-old, even a bright one, is nothing like so extensive as that expected of a teenager or adult, thus enabling a 4-year-old with a large recognition vocabulary (usually all that is required) to score at a high percentile for his or her age. This advantage disappears over time, because testing when the child is older will require more subtle language comprehension and usage, and many more language tasks will enter into the assessment..

          I’ve taught some, and tested many others, who displayed this discrepancy between cognitive ability and vocabulary at age 5-8. While most were in the average range with high vocabulary subscores, some were actually in the developmental disability range (IQ around 50-65), one with a vocabulary at the 88th percentile.

          Average or even “mentally retarded” preschool and K children can learn enough vocabulary — through a lot of explicit interaction and direct instruction (not flashcards or dictionaries) to outperform their actual cognitive level by several SD. Usually, of course, this does not lead them to a “gifted” classification but it can, if coupled with a few other strong performances in subtests, make them appear stronger than they actually are. Paired-associate learning can be very easy for some young children and they can learn a great deal of “rote” material.

          If you take a child that is reasonably bright and coach them on areas of the WPPSI (the test usually used with that age group) such as information, receptive vocabulary, picture naming, picture concepts, and/or subtests that are known to yield practice effects, such as coding, symbol search and matrix reasoning, this could certainly affect their score — much more than similar “coaching” would assist an older person, because the range for 4-year-olds is much narrower. Subtests like similarities are much less amenable to improvement this way, although this too can be systematically taught (not enough to make a very low-ability child appear gifted however).

          In the past it has not been a big concern that much effort would be put into “teaching to the test” where preschoolers are concerned but it has become more of an issue in recent years, especially in the light of some notorious cases. Reputable trained psychometricians (I’ll include myself in that category) would never use our knowledge of the test items and processes to deliberately subvert the integrity of the assessments, but we know that some have and do.

          Cognitive assessment of children that young is considered provisional at best. IIRC, about a third of children considered “gifted” at age 5 end up as adolescents in the average or superior (but not gifted) range, and the opposite can also be true. Measured IQ is not stable in this age .

          Another consideration, in the case of particular tests, is whether there is a low ceiling or not. The Weschler tests have age-based ceilings beyond which a child cannot score, while the Stanford-Benet is normed on a much wider age sample, and this yields different results, especially at the extremes (low and high).

  2. Test prep – for four year olds? This is criminally insane. How are all these idiots in education able to put forth such insane policies without being censured and arrested??

    • You’re nuts. You shouldn’t be allowed on the Internet. What do you think is an arrestable offense? Anything you don’t agree with? Dear Leader in North Korea would agree with you – but then again, if you were there you’d be dead in a concentration camp pretty quickly! Liberalism is a mental disorder.

  3. These stories make me sad. So much energy devoted to denying children educational opportunities. How much does it cost the city just to run the “qualification” circus?

    Any system which can be gamed will be. Especially in New York. The article states, “Mr. Davison has suggested that the schools should develop their own test that would be administered by the schools themselves, and not by psychologists, who are widely believed to be, along with professors and consultants, among those supplying the tests to test prep companies.”

    From the descriptions of the tests in the NYT and New York Magazine, I don’t think they end up with many gifted children. They end up with many prepped children, in my opinion. Why not allow parents to sign their children up for more demanding academics? Just do away with the fiction the process finds the “gifted.” Whether a child is naturally very quick, or can learn from prep, it should not be the business of the state to dub some winners and others losers at the age of 4.