Nonverbal Reasoning
What the nonverbal battery measures
The Nonverbal battery uses geometric shapes and figures rather than words or numbers, so it measures reasoning about form, position, and transformation. At Levels 13–18 the three subtests are Figure Matrices, Paper Folding, and Figure Classification. The core habit is analytic: break each figure into separate attributes and track them one at a time.
Figure Matrices
A 2×2 or 3×3 grid with one cell missing. Read across each row and down each column, and confirm that the rule you find holds in both directions. Attributes to track: shape type, number of sides or elements, size, shading and fill, rotation (angle and direction), reflection, position, and the addition or removal of parts. In 3×3 grids a rule may apply across rows only, or across the whole grid.
Paper Folding
Holes are punched through folded paper and you choose the unfolded pattern. Each fold doubles the number of layers, so one punch through n layers makes n holes. Each fold line acts as a mirror: every hole reappears at the same distance on the other side of that line, with mirrored orientation. Holes punched exactly on a fold line merge into one larger hole, so the count drops.
Figure Classification
Three figures share a property; choose the fourth that shares it (or, in some forms, the one that does not). State the rule so that it covers all three given figures, then confirm it excludes every distractor.
Worked example
A square is folded in half top-to-bottom, then in half left-to-right, giving four layers. One hole is punched through all layers near the top-left corner. How many holes appear when unfolded, and where? Layers: 2 × 2 = 4, so four holes. Positions: undoing the second fold mirrors the punch across the vertical fold line; unfolding the first fold mirrors both holes across the horizontal fold line. The result is four holes arranged in a rectangle, each the same distance from the two fold lines as the original punch.
Traps that cost marks
- Counting shapes instead of layers in paper folding, or forgetting a fold.
- Using translation where a fold produces reflection.
- Over-fitting an attribute rule from one figure instead of all three.
- Missing rotation direction (clockwise versus counterclockwise) or rotation amount.
- Assuming that a rotating figure also changes shading - track each attribute independently.