Accuracy before compression
Write or say a repeatable method during untimed review. Speed up the method—not a guess.
Free field guide · updated for 2026
Thorough directions, repeatable checks, pacing guidance, and a study system for the full DAT. Official rules are separated from DAT TRAIN recommendations so you always know which is which.
BEFORE YOU TRAIN
The official section contains six spatial-reasoning families. A section-wide average is 40 seconds per item, but forcing every item into exactly 40 seconds is poor pacing: some visual comparisons should be fast, while multi-step folds or views deserve more deliberate work.
Apertures, Paper Folding, and Cube Counting use five choices. View Recognition, Angle Discrimination, and Spatial Relations use four. DAT TRAIN preserves those family-specific counts.
The guide teaches the complete official rule set. Apertures rotate among five connected-solid families—core-plus, stair, arch, cross, and fork—with multiple axis permutations and non-principal entry orientations. Four families deliberately omit the former 2×2×2 core, and each item derives its conservative all-rotation rejection threshold from the largest certified occupied cube inside that solid. TFE now uses a separate exact-voxel engine for connected bridge, cantilever, and offset-frame solids. Every generated object contains a genuine elevated gap or overhang that cannot be represented as floor-filled height columns. A target-specific slice certificate proves that the two supplied silhouettes reconstruct the missing top, front, or end drawing uniquely; visible and hidden edges are then derived from that same solid. Paper Folding mixes equal, unequal, and inward diagonal corner folds across two- and three-step items, with all 64 original paper cells tracked exactly through every reflection. Curved or free-angle creases remain outside the current generator. Spatial Relations now alternates triangular-prism and tetrahedron nets across eight rotation/reflection layouts. Each face carries a unique color-symbol identity and a corner-anchored arrow. The solver verifies closed polyhedron edges, exact face incidence at every vertex, and both the faces and marker directions in every answer option. These limits remain visible until broader generators pass the same correctness gates.
Write or say a repeatable method during untimed review. Speed up the method—not a guess.
Name the violated dimension, crease, adjacency, or face count. “Looks wrong” is not reviewable.
Knowledge, visualization, diagram-reading, and pacing errors need different remedies.
THE PAT PLAYBOOK
Each card begins with the official task and rules, followed by DAT TRAIN’s recommended solving routine.
Apertures
Choose the one exact opening through which the object can pass completely.
The object may be turned before entry, including toward a side not shown. After entry begins, it cannot turn or twist. The object and openings use a common scale, and only one aperture is correct.
Use principal views to build confidence, then train multi-axis entry orientations and varied connected solids while checking the complete silhouette at one fixed scale.
View Recognition
Infer the missing orthographic view from the other two views.
Top is always upper left, front lower left, and end lower right. Views have no perspective. Hidden edges are dotted. The missing drawing may be top, front, or end.
Begin with floor-supported stacks, then train elevated beams and gaps. Explain one eliminated option aloud by naming the exact width, depth, height, support, or visibility conflict.
Angle Discrimination
Rank four angles from smallest to largest.
The ray length, placement, and orientation do not change the angle. Judge only the opening between the rays.
Train accuracy without a timer first, then use short bursts so speed grows from a stable visual comparison.
Paper Folding
Predict every hole after the folded paper is completely unfolded.
Broken lines show the original paper; solid lines show the retained folded paper. The sheet is never turned or twisted and remains inside the original square. Multiple folds precede one punch.
Draw one crease at a time during review and label each reflected pair before viewing the explanation.
Cube Counting
Count cubes with a requested number of painted faces.
Equal cubes are cemented together, then painted everywhere except the bottom resting surface. Hidden cubes exist only when needed to support cubes above. Zero is never the keyed answer.
Keep the full 1–5 tally even when only one category is asked; the total becomes an error detector.
Spatial Relations
Choose the solid that can be produced from the flat pattern.
Folding preserves face adjacency, edge connections, and the orientation relationship among markings. Faces cannot overlap or pass through one another.
During review, name the three faces at one vertex and state whether each marker points toward or away from it before checking the key.
PACING WITHOUT PANIC
A planning number, not a rule for every question.
Use the block boundaries as coarse checkpoints.
Protect later questions from one stubborn diagram.
FULL DAT STRATEGY
A good plan changes according to evidence from your attempts. It is not a fixed number of videos, chapters, or hours.
Total administration time is 5 hours 15 minutes. The scheduled break comes after Natural Sciences and PAT.
Start with the current ADA specification. Turn every listed domain into a checklist so familiar topics do not hide untested gaps.
Close the notes and answer from memory. Practice is a learning event, not merely a score report.
Review immediately enough to repair the reasoning. Record the cause: missing fact, wrong method, diagram error, or pacing decision.
Return to misses after time has passed. Increase the interval when recall is strong and shorten it when the method is still unstable.
After a short blocked foundation, mix confusable topics so the question—not the worksheet heading—must tell you which method to use.
Add strict section timing only after the underlying method is dependable. Full forms train endurance and transitions; they do not replace review.
FLEXIBLE SIX-STAGE TEMPLATE
Move forward when the evidence supports it; keep weak objectives in spaced review while new material enters.
Chose formula before identifying the requested quantity.
Write givens, target, units, then choose relationship.
Tomorrow, then 4 daysReflected across fold 1 before undoing fold 2.
Unfold strictly in reverse order; mirror every current hole.
Next short PAT setTEST-DAY CHECKLIST
Official DAT scores are psychometrically equated and reported from 200 to 600 in 10-point increments. A raw practice percentage is useful for diagnosing objectives and pacing, but it is not an official score estimate.
SOURCE DESK
Official rules come from the ADA. Study-system recommendations are informed by experimental learning research and are labeled as recommendations, not guarantees.
Official scope, timing, scoring, rules, and administration policies.
OPEN SOURCE ↗The controlling directions and examples for all six PAT families.
OPEN SOURCE ↗Current notices, official practice links, the Prometric tutorial, and program updates.
OPEN SOURCE ↗Roediger & Karpicke: testing improved delayed retention relative to repeated study.
OPEN SOURCE ↗Cepeda and colleagues: the useful study gap depends on how long knowledge must be retained.
OPEN SOURCE ↗Butler & Roediger: feedback improved later accuracy and reduced learning of multiple-choice lures.