Math Competitions

    NGAT vs NNAT: Why the NNAT’s Own Inventor Built a New Test

    North Star Contests Team

    NGAT vs NNAT: the NNAT's own inventor built the NGAT and gave it a verbal section made of pictures. What that means for your child's gifted screening.

    Jack Naglieri spent twenty-five years making one argument. If you want to find gifted children fairly, take the words out of the test.

    Then he built a new test and gave it a verbal section.

    Think about what that admission cost him. Naglieri built his whole career on the idea that a fair test uses no words. Nobody had more to lose by saying that idea was incomplete.

    He said it anyway. Then he built the replacement.

    That one fact tells you more than any comparison chart will. So let's walk through what changed, what didn't, and what it means if your child is about to be screened.

    The short version

    If you're in a hurry:

    • NNAT3 (Naglieri Nonverbal Ability Test, 3rd Edition, from Pearson) is one test. No words at all. Just shapes in a grid. 48 questions, 30 minutes.
    • NGAT (Naglieri General Ability Tests, from MHS) is three separate tests: Verbal, Nonverbal, and Quantitative. A school can give one, two, or all three. About 30 to 35 minutes each.
    • Same psychologist designed both. Both try to measure how your child thinks, not what your child has been taught.
    • The NGAT is the better-designed test. The NNAT has twenty-five years of independent research behind it. Some of that research is not kind.
    • You almost certainly don't get to choose. Your school board chooses.

    That last point matters more than parents expect. We'll come back to it.

    What the NNAT actually is

    The NNAT is a pattern test. Your child sees a grid of shapes with one piece missing. They pick the option that completes it.

    Shapes rotate. Colours change. Pieces get added or taken away. Patterns march across the grid.

    There are no words anywhere. Not in the questions. Not even in the instructions, which are given in pictures. A child who arrived in the country three weeks ago sits exactly the same test as a child born down the street.

    Here's the structure:

    • 48 questions, 30 minutes of testing (35 to 45 minutes once you count instructions and settling in)
    • Seven levels, A through G, covering kindergarten through Grade 12
    • Available on paper or online
    • One score, the Naglieri Ability Index (NAI). Average is 100. Your child is compared only to children of almost exactly the same age. At the youngest levels, that group is a one-month age band

    Pearson built the NNAT3's scoring using about 10,000 students during the 2014–15 school year, across 113 schools in 27 U.S. states.

    It's a clever test. It is also, by design, a test that looks at one thing.

    What the NGAT actually is

    Here's the part most guides get wrong.

    The NGAT's Verbal test has no words in it.

    Read that again. A student sees six pictures. Five of them share an idea. One doesn't. Find the odd one out.

    Think of it this way. You could ask a child to define the word "vehicle." Or you could show them a car, a bus, a bicycle, a train, a canoe, and a sandwich.

    The first version tests whether they know an English word. The second tests whether they understand the idea.

    The three tests work like this:

    • Naglieri–Verbal: six pictures. Five share an idea. Pick the one that doesn't belong.
    • Naglieri–Nonverbal: grids of shapes. Work out the pattern from sequence, colour, and position. This is the NNAT's old territory.
    • Naglieri–Quantitative: numbers and shapes in a pattern, using basic math ideas. No word problems.

    The rest of the design is built to remove anything that could trip a child up for the wrong reason:

    • Online only. There is no paper version
    • Animated video instructions, so no child fails because they misheard a teacher
    • Three practice questions before each test
    • About 30 to 35 minutes per test. A student can sit more than one on the same day, in any order
    • Grades K–9, with national scoring for K–6 and local scoring through Grade 9
    • Results come back as one score, a percentile, and a stanine (a 1-to-9 band). There are no subscores

    So the NGAT looks at three things and reports one number. That tells you what it's for. It isn't trying to sort your child into a type. It's trying not to miss them.

    The real difference: one lens or three

    Strip away the marketing and this is the whole debate.

    The NNAT says: words are where unfairness sneaks in, so remove the words.

    The NGAT says something different. Reasoning with words and reasoning with numbers are real parts of thinking. Don't delete them. Just stop asking about them in a language a child hasn't learned yet.

    Judging a gifted child on a nonverbal test alone is like judging a triathlete by watching only the swim. You'll learn something real. You'll also miss two thirds of the athlete.

    And here's why that matters for your family. A child can be remarkable at one kind of thinking and merely good at the others. A one-lens test either catches that child or it doesn't. A three-lens test gives them three chances to be seen.

    The uncomfortable evidence

    The NNAT was sold on fairness. So it's fair to ask what researchers found once they tested that claim. Especially researchers who didn't build it.

    In 2008, David Lohman and his colleagues gave three tests to 1,198 elementary students. About 40% were English-language learners. They compared the Raven, the NNAT, and the CogAT's nonverbal section.

    Three things stood out:

    1. Nonverbal did not mean equal. English-language learners averaged 90.7 on the NNAT. Other students averaged 100.7. That's a ten-point gap on a test built to close exactly that gap.
    2. The scores were shaky. At the high cutoffs gifted programs actually use, the NNAT's margin of error was about double the CogAT's. Roughly 14 points wide, against about 6.
    3. The youngest scores were inflated. At Level A, 3.4 times more children scored 130 or above than the math says should have.

    Their conclusion was blunt. Nonverbal tests do not level the playing field.

    Taking the words off the page doesn't take the disadvantage out of the childhood.

    Then, in 2021, another team reviewed 1,714 studies and analysed the 29 that met their standards. They found the NNAT did identify more students of colour than other nonverbal tests. They also found those students were still underrepresented in gifted programs.

    And they noticed something else. The results looked better in studies where Naglieri himself was one of the authors.

    Hold on to that. The main fairness study on the NGAT is a 2024 paper in Psychology in the Schools. It reports few differences by race, ethnicity, gender, or parent education. Naglieri is one of its authors too.

    That doesn't make it wrong. It makes it early. Nobody has spent twenty years trying to poke holes in the NGAT yet.

    The NGAT has earned the benefit of the doubt. It hasn't yet earned the benefit of the evidence.

    The scoring story nobody tells parents

    In 2018, Pearson quietly rescored the NNAT3's lower levels using a bigger sample of students.

    Average scores dropped by about 4.5 points.

    Nothing about the test changed. Not one question. Not one instruction. The same child, giving the same answers, now scored four and a half points lower. All because the comparison group had changed.

    This is the most useful thing a parent can understand about these tests, so here it is as plainly as we can say it:

    A test score is not a height. It's a place in a line.

    Your child's 96th percentile doesn't describe your child. It describes where your child stands next to whoever the publisher lined them up against. Change the line, change the number.

    Which raises a question worth asking your board, especially in Canada. Whose line?

    The NGAT's national scoring is built to match U.S. census data. That's a sensible thing for a U.S. publisher to do. It also means a Canadian board using national scoring is comparing your child to a group of American kids.

    This is exactly why local scoring exists, and why boards should use it. The CCAT-7, for what it's worth, is scored against Canadian students.

    Nobody puts that paragraph in the letter that comes home.

    If you're in Canada

    Here's what most comparison articles bury. Most Canadian children will never sit an NNAT. It's overwhelmingly a U.S. test. New York City used it as half of its gifted exam until the city scrapped that exam in 2021.

    In Ontario, the real comparison isn't NGAT versus NNAT. It's NGAT versus CCAT-7.

    The Peel District School Board now screens most Grade 4 students with the NGAT. Every child's results go to a Student Review Meeting. That meeting isn't gated by a score.

    From there, Peel says some students who scored at the 96th percentile would be considered for identification by an IPRC. That's the committee meeting where a board formally identifies a child as gifted. It usually involves a full one-on-one assessment with a psychologist first, and identification generally requires results about two standard deviations above average.

    Notice the word some. The percentile opens a door. It doesn't walk your child through it.

    Most other Ontario boards still use the CCAT-7. The TDSB is one of them, and it screens a year earlier, in Grade 3.

    If all of this sounds confusing, that's because it is. We wrote a full board-by-board walkthrough here: Two of Ontario's Biggest School Boards Quietly Changed the Gifted Test. It covers the four stages, the cutoffs each board publishes, and your rights under Regulation 181/98, including appeal deadlines almost nobody uses.

    If you're in the US

    Your district might use either test, plus a dozen others.

    The NNAT is the one that's been around for years. The NGAT is the newcomer, and it's picking up districts.

    Don't ask your district office "which test?" Ask these three questions instead:

    • Which tests, and which sections? A district can give one NGAT test or all three. That completely changes what your child is measured on.
    • National scoring or local scoring? This decides whether your child is ranked against the whole country or against the school down the road. A district that doesn't look like the national average can identify a very different number of kids depending on which it picks.
    • What's the cutoff, and is it the only way in? These tests aren't precise enough to split hairs. If there's one hard cutoff and your child lands near it, the result is close to a coin flip.

    So which test is better?

    On design, the NGAT. Three lenses beat one. Testing word reasoning with pictures is a smarter fix than deleting word reasoning altogether. Video instructions and built-in practice questions remove problems that have nothing to do with ability.

    On evidence, the NNAT, for now. Only because it's been studied longer, including by people trying to prove it wrong. That's not praise. It's just what happens when a test has been around a while.

    But honestly, this is a bit like asking whether a Phillips or a flathead is the better screwdriver. Fair isn't the same as accurate. Neither one is the same as useful.

    Both of these are screening tests. Neither is meant to be the final word on your child. In Ontario, neither one is. Real identification runs through a full assessment and a committee, not a 30-minute Tuesday morning.

    What you should actually do

    School boards will tell you no preparation is needed. Peel's letter to families says it outright: "There is no preparation necessary for the NGAT."

    They're mostly right, and it's worth being honest about why.

    You can't cram for a reasoning test. There's nothing to memorize. No formula sheet. No vocabulary list. The research on coaching is thin and inconsistent, and any tutor promising you a percentile jump is selling you something.

    But "you can't study for it" and "nothing you do matters" are two different sentences.

    What actually helps is short and boring:

    • Remove the surprise. A child who has never seen a shape-pattern question spends the first four questions working out the rules instead of answering. On a 48-question, 30-minute test, that's expensive.
    • Sleep and breakfast. Every board says this because it's annoyingly the most powerful thing on the list.
    • Say out loud that it doesn't decide anything. Kids pick up your nerves fast. Halton-area experts have warned about the stress these screenings put on children and families, and a lot of it comes from simple misunderstandings about what the test is.
    • Keep the thinking muscle warm all year. Not test prep. Just puzzles, patterns, and problems that don't come with an obvious method.

    That last one is the honest case for what we do at North Star. A child who wrestles with hard problems every week isn't "prepped" for a screening test. They're just used to the feeling of not knowing the answer yet.

    That feeling is most of the test.

    If your child is being screened this year, you can find out where they actually stand in about 15 minutes. Our free NGAT-style assessment is built for Grade 4 and comes back with a gap report. No card needed.

    The same page carries the Full Test Prep Pack: close to 2,000 questions across all three NGAT styles, from picture classification and picture analogies to number series, number matrices, figure matrices, pattern completion and spatial visualization. It's $59 CAD one-time with three months of access, and it's included in the Season Pass.

    If your board still uses the CCAT-7, start with CCAT-7 practice instead. It's the same pack, with free starter questions set up by board for Toronto, York Region and Durham.

    And if you just want contest practice rather than screening prep, our past contest papers are free, with step-by-step solutions.

    And if you're still deciding whether competition math is worth your family's time, start with our roundups of Caribou contest alternatives and K-12 math scholarships in the US and Canada.

    One last thing

    Whichever test your board uses, remember what it's really doing.

    It is spending half an hour guessing at something that takes a whole childhood to show. Sometimes it guesses right. Sometimes a bright kid has a bad Tuesday. Sometimes a quiet kid finally gets noticed. And sometimes the scoring shifts under everyone's feet and 4.5 points vanish.

    The NGAT is a better guess than the NNAT. It's still a guess.

    Your job isn't to beat the test. Your job is to make sure one number never becomes the story your child tells themselves about how smart they are.


    Screening practices change from year to year. Always confirm the current test, grade level, and cutoffs with your own school board's special education office before making decisions.

    Sources

    More from Math Competitions