Prepare for the NNAT® format, not a pile of answer keys.

The NNAT®3 gives a child 48 questions in about 30 minutes — roughly 37 seconds per question on average. That pacing is a useful anchor for preparation: a child doesn't need to be fast so much as unsurprised. Time spent figuring out an unfamiliar layout on test day is time that isn't going toward the actual reasoning, so format familiarity is worth building deliberately before test week arrives.

Start by learning the task

Before increasing speed or volume, make sure the child understands what each format is asking. For analogy questions, name what changes and what stays the same. For matrices, inspect both rows and columns. For Pattern Completion, compare all four cutout edges rather than the most obvious feature in the center.

What children actually see

Four question types, each with its own logic.

These are real GateBuddy bank examples, one from each NNAT® format, so families can see the kind of visual reasoning the test actually asks for.

01

Pattern Completion

Complete the window

Follow lines, shapes, and colors through a missing section of one continuous design.

02

Reasoning by Analogy

Apply the change

Discover how one picture changes, then use the same relationship again.

03

Serial Reasoning

Finish both patterns

Read changes across rows and columns to determine the missing picture.

04

Spatial Visualization

Picture the result

Combine, rotate, or fold visual information in your mind to find the matching answer.

This stage is worth slowing down for, even though it can feel like the least productive part of preparation. A child who can correctly answer a question without being able to say why the answer is correct has often pattern-matched rather than reasoned through it, and that distinction tends to show up later on a fresh question that looks similar but isn't. Asking a child to explain their answer out loud, even briefly, is a fast way to tell which is happening.

Use short, focused sessions

Five careful questions with real explanations can teach more than a long packet finished without feedback. Short sessions make it easier to identify a specific misconception, review it, and try a fresh example while the rule is still clear in the child's mind. A session doesn't need to end because time ran out — it can end as soon as the goal for that day (one format, one skill, one specific mistake pattern) has actually been addressed.

Make explanations do real work

An explanation should not merely describe the pictures. It should state the transformation or organizing rule, identify the features that remain constant, and show how the correct answer follows from that rule — not just confirm which choice was right. A weak explanation says "the answer is C because it matches"; a useful one says "the shape rotated 90 degrees and stayed the same color, so C is the only choice that does both."

Build from one clear change toward several at once

Within any of the four formats, the easiest questions involve a single visual change — one shape gets bigger, one color swaps, one piece rotates. The harder questions ask a child to track two or three changes happening at the same time, or to satisfy a rule that has to hold in more than one direction, like a Serial Reasoning matrix where the row rule and the column rule are both active. Starting with single-change questions isn't about making practice artificially easy; it's about letting a child learn to identify one transformation cleanly before asking them to track several simultaneously. Moving to multi-change questions too early tends to produce guessing rather than reasoning, since the child hasn't yet built the habit of isolating what actually changed.

Add full-length practice deliberately

A complete practice test, timed to the real 30-minute window, helps a child experience sustained attention, question navigation, and time awareness in a way that short drills don't. It should come after basic format familiarity, not replace it — a child's first encounter with all four formats shouldn't also be their first timed test. A full test also surfaces something short sessions can't: how a child's accuracy holds up in question 40 compared to question 5, which is a different skill than getting any single question right.

Review patterns in mistakes

One wrong answer is a moment. Several mistakes involving reflection, edge continuity, or overlay point to a skill worth revisiting directly. The most useful mistake review shows the exact questions missed and groups them by the reasoning skill behind them, not just a raw score.

What to say (and what to avoid) during practice

The words a parent uses during a practice session shape whether a child treats a wrong answer as useful information or as a small failure. Prompts that name a specific thing to look at, rather than a general instruction to try harder, tend to work best:

• "What changed?" and "What stayed the same?" — for Reasoning by Analogy
• "Does your rule work going across the row and down the column?" — for Serial Reasoning
• "Did anything turn or flip?" — for Spatial Visualization
• "Take one more look at the whole pattern before you choose." — for any format

Phrases worth avoiding include "you know this," "that was easy," and "why are you rushing?" — none of them point to what to do differently, and all of them attach a small amount of judgment to a child who is still learning the format. It's also worth resisting the urge to name the rule before a child has had a real chance to find it themselves; the reasoning practice happens in the searching, not in being told the answer.

How much is too much?

Signs that a session has run past its useful point include rushing through remaining questions without reading them, guessing without looking at all five choices, or visible frustration that wasn't there ten minutes earlier. At that point, stopping and returning to the format on a different day is more productive than pushing through — a tired child practicing a new visual format is more likely to build a wrong habit than a right one. There's no fixed session length that works for every child; the signal to stop is behavioral, not a timer.

A simple sequence to follow

1. Try a short sampler across all four formats.
2. Practice the least familiar format at a foundation level.
3. Review missed questions with rule-based explanations.
4. Move to developing and challenge questions gradually.
5. Take a complete timed test.
6. Use the result to choose the next focused session.