Nonverbal Battery: Figure Matrices, Paper Folding, and Figure Classification
The Nonverbal battery of CogAT Level 11 uses shapes and figures only, with no reading and no arithmetic. Its three subtests, Figure Matrices, Paper Folding and Figure Classification, measure inductive reasoning: spotting a rule in a visual pattern and carrying it to a new case.
Figure Matrices
A 2 by 2 grid shows a change in the top row and asks you to repeat it in the bottom row. State the change as a rule: rotated 90 degrees clockwise, shrunk, shading reversed, one side added, the dot moved to the opposite corner. Check the same rule down the columns, because the item is built so that it holds both ways; if your row rule contradicts the column rule, it is the wrong rule.
Paper Folding
A sheet is folded, a hole is punched, and you choose the unfolded sheet. Two rules cover most items. First, count: each fold doubles the number of holes, so one fold gives 2 and two folds give 4, unless the punch sits exactly on a fold line, where half-holes join into a single hole. Second, position: every hole is mirrored across every fold line, including folds made earlier.
Figure Classification
Three figures share one attribute and the answer must have it too. Count before you choose: sides, vertices, enclosed regions, lines of symmetry, straight versus curved lines, shaded versus white areas.
Worked example
Fold a square in half left to right, then top to bottom. Punch one hole in the middle of the folded packet, away from all edges. Unfold. Each fold doubles the holes, so 1 x 2 x 2 = 4 holes, one in each quadrant; each is the mirror image of the punch across the vertical fold line and then across the horizontal one. Eliminate any option with 2 or 3 holes, holes sitting on the fold lines, or four holes with no mirror symmetry.
Traps that cost marks
- Getting the right number of holes but the wrong positions.
- Forgetting that a second fold also mirrors the holes produced by the first.
- Treating a punch on a folded edge as a full hole, when it unfolds into one whole shape rather than two.
- In matrices, changing the wrong attribute, such as resizing when the row changes rotation.
- In classification, choosing a figure that matches the look of the set but shares the counted feature with only two of the three.