7.2 Figure Classification: Spatial Commonality & Rule Abstraction

7.2 Figure Classification: Spatial Commonality & Rule Abstraction

Quick Answer: The Figure Classification subtest on CogAT Level 9 (Grade 3) contains 20 items to be completed in 10 minutes (30 seconds per item). Each question displays three target geometric figures that belong together because they share a hidden spatial rule. Students must select the figure from five answer choices that belongs in the same category. Mastery requires pure visual inductive reasoning ($G_f$), scanning across four key rule domains (topological, geometric, symmetry, and internal composition), and applying the Goldilocks Rule to identify a principle that is neither too broad nor too narrow.

---

Subtest Architecture & Rapid Pacing Realities

Figure Classification represents the third and concluding subtest of the CogAT Nonverbal Battery. At Level 9 it has 20 questions in 10 minutes:

With 30 seconds per question, Figure Classification does not allow time for complicated scratch-paper sketches or prolonged deliberation. Third graders must develop rapid perceptual filtering—instantly scanning the three target figures, isolating the shared structural property, and testing that rule against the five choices.

---

Inductive Visual Reasoning: Abstracting Rules Without Words

In cognitive psychology, inductive reasoning is the process of inferring a general principle from a set of specific observations:

$\text{Target 1 (Specific)} + \text{Target 2 (Specific)} + \text{Target 3 (Specific)} \implies \text{Abstract Category Rule}$

+------------------------------------+       +------------------------------------+
|          TARGET FIGURES            |       |           ANSWER CHOICES           |
|  [ Target 1 ] [ Target 2 ] [ Target 3 ]  | ===>  |  [ Option 1 ]      [ Option 2 ]    |
|       (Share Hidden Property)      |       |  [ Option 3 ] [ Option 4 ] [ 5 ]  |
+------------------------------------+       +------------------------------------+
                                                        |
                                                        v
                                             Select the ONLY figure
                                             that obeys the Rule!

The Absence of Language and Arithmetic

Unlike Verbal Classification (which tests word concepts like apple, banana, orange) or Number Puzzles (which tests arithmetic balance), Figure Classification operates in a purely visual-spatial universe:

• There are no words, definitions, or verbal labels provided.
• There are no mathematical equations or numerical quantities given.
• The student must translate raw visual features into an abstract structural relationship.

---

The Four Core Classification Rule Domains at Level 9

Riverside does not publish a list of rule types, but practice items draw on four broad domains. Riverside's own Level 9 practice items use rules such as "each has a line inside that slants to the left" and "each has a square and a triangle next to each other," which belong to the internal-composition domain below. Train students to cycle through all four domains:

1. Topological Properties

Topological rules focus on continuous spatial qualities that do not depend on exact measurements:

• Closed vs. Open Curves: Do all target figures enclose a distinct interior region (polygons, ovals, closed loops), or do they possess free endpoints that leave the interior open (spirals, horseshoe curves, zig-zag lines)?
• Intersecting vs. Non-Intersecting Lines: Do line segments cross over one another (like a figure-eight, an X, or overlapping loops), or do they maintain a simple perimeter without self-intersection?
• Number of Disconnected Components: Is each figure a single connected drawing, or does it consist of two, three, or four separate, detached elements?

2. Geometric & Polygon Properties

Geometric rules focus on precise Euclidean attributes:

• Number of Sides and Vertices: Are all three target shapes triangles (3 sides), quadrilaterals (4 sides), pentagons (5 sides), or hexagons (6 sides)? Alternatively, do all targets have an even number of sides (4, 6, 8) or an odd number of sides (3, 5, 7)?
• Presence of Right Angles ($90^\circ$): Does every target figure include at least one perpendicular, square corner ($90^\circ$ angle)?
• Parallel Line Segments: Does every figure contain at least one pair of parallel edges (such as trapezoids, parallelograms, or rectangles)?
• Curved vs. Straight Boundaries: Are the figures formed exclusively from straight line segments, exclusively from smooth curves, or from a specific combination (e.g., exactly two straight sides and one curved side)?

3. Symmetry & Chirality

Symmetry rules evaluate balance across reflectional and rotational axes:

• Bilateral (Reflective) Symmetry: Does every target shape possess at least one line of symmetry where it can be folded into two matching halves? Notice whether the axis is strictly vertical, strictly horizontal, or diagonal.
• Rotational Symmetry: Does the figure look identical when rotated around its center point by $90^\circ$ or $180^\circ$ (like an S-curve, pinwheel, or propeller)?
• Strict Asymmetry: Are all three figures intentionally scalene or irregular, possessing zero axes of symmetry?

4. Internal Composition & Structural Nesting

Compositional rules govern how multiple shapes interact inside a single frame:

• Internal Partitions: Is each target figure divided into equal fractional sectors (e.g., bisected into halves, or quartered into fourths)?
• Concentric and Nested Relationships: Does a large outer container enclose a smaller inner shape? Is the inner shape always identical to the outer container (a small square inside a large square), or is it always different?
• Shading and Pattern Contrast: Are all figures filled with a specific shading state (solid black, striped, hollow white)? In composite figures, does the smaller inner shape always contrast with the outer shape (e.g., black inner element inside white outer element)?
• Quantitative Attribute Correspondence: Does the number of internal dots, stars, or lines equal the number of sides on the outer container (e.g., 3 dots in a triangle, 4 dots in a square, 5 dots in a pentagon)?

---

Common Spatial Rules Taxonomy Table

---

The Goldilocks Rule for Spatial Classification

The most sophisticated challenge on Figure Classification is selecting the correct level of rule specificity. A student's hypothesized rule can easily be either "too broad" or "too narrow."

The Goldilocks Rule dictates that the valid classification rule must be just right: it must be specific enough to unite all three target figures while matching exactly one answer choice.

       [ TOO BROAD RULE ]        --->  Matches all 3 targets, but also several answer choices!
       "They are all closed shapes."   (Fails to discriminate)

       [ TOO NARROW RULE ]       --->  Matches only 1 or 2 targets; excludes valid candidates!
       "They are all equilateral triangles." (Overlooks abstract commonality)

       [ "JUST RIGHT" RULE ]     --->  Unites ALL 3 targets and matches EXACTLY ONE answer choice!
       "Polygons with exactly one vertical line of symmetry." (The Goldilocks Solution)

Calibration Walkthrough

Imagine an original practice item with these three target figures:

• Target 1: A circle with one line inside that slants up to the right.
• Target 2: A square with one line inside that slants up to the right.
• Target 3: A triangle with one line inside that slants up to the right.

The five answer choices are a hexagon with no line, a star with a horizontal line inside, a pentagon with a line slanting up to the left, an oval with a line slanting up to the right, and a diamond with two crossing lines.

Calibrate the rule:

• Too broad: "They are all closed shapes." Every choice is a closed shape, so the rule cannot pick one.
• Still too broad: "They all have a line inside." Four choices (the star, the pentagon, the oval, and the diamond) have lines inside.
• Too narrow: "They all have four sides." That rule fails the targets themselves, because the circle and the triangle do not have four sides.
• Just right: "Each has one line inside that slants up to the right." Only the oval fits.

This mirrors the advice in Riverside's practice guide: when more than one choice fits your first rule, go back to the three figures and look for a more precise rule. Its guide also lists the mistakes to avoid: overlooking a critical feature (noticing the outer shape but ignoring the shading), choosing an answer that matches only one of the three figures, and choosing before checking every answer choice.

---

The 4-Step Inductive Abstraction Protocol

During the rapid 27-second window for each item, third graders should execute this streamlined protocol:

[ Step 1: Scan Targets ]      --> Compare Targets 1, 2, and 3 across the 4 core domains.
            |
[ Step 2: Formulate Rule ]    --> State the shared property aloud in a concise phrase.
            |
[ Step 3: Apply Goldilocks ]  --> Verify the rule is neither too broad nor too narrow.
            |
[ Step 4: Scan Choices ]      --> Find the single option that satisfies the exact rule.

1. Step 1: Rapid Scan across Domains. Check topology first (closed vs. open). If all are closed, count sides and corners. Check for right angles, parallel sides, and symmetry.
2. Step 2: Formulate the Rule Sentence. Articulate the rule clearly (e.g., "Each figure has four sides and exactly one right angle").
3. Step 3: Goldilocks Check. Does the rule genuinely unite all three target figures without exception?
4. Step 4: Filter Options & Lock Answer. Inspect all five choices. Four will violate the rule; eliminate them and select the single matching figure.

By training on this systematic classification taxonomy, students replace random guessing with disciplined visual deduction, securing high scores on this fast-paced subtest.