Visual Spatialization Ability Test
30 mental-rotation trials in about 8–10 minutes. Each trial shows a reference cube, then a cued rotation (for example, 135° clockwise), then five candidate cubes — choose the one rotated exactly as cued. 2 points per correct choice, maximum 60. Difficulty rises across three blocks. Free with a registered account.
The three blocks
Trials 1–10
Rotation cues between 45° and 180°, with widely spaced answer options. A gentle on-ramp to the response format: fixation cross, reference cube, rotation cue, then the five-option answer display.
Trials 11–20
Rotation cues widen to 30°–240° and the incorrect options sit closer to the target orientation, so a rough impression of the rotation is no longer enough.
Trials 21–30
The hardest block: cues span 20°–320° and the distractor orientations crowd tightly around the correct answer, demanding a precise analog rotation of the mental image.
How it works
What it measures
Visual-spatial mental rotation ability — the analog transformation of a visual image — in the tradition of Shepard & Metzler (1971) and the Vandenberg & Kuse (1978) Mental Rotations Test, using an original deterministic item bank presented identically to every respondent. Mental rotation is a well-studied component of nonverbal, visually mediated reasoning: the task involves holding a shape in mind and transforming it without any verbal content, which is one facet of the broader family of nonverbal learning processes. A single 30-trial task samples that facet only; it is an indicator of current performance, not a comprehensive measure of nonverbal ability, and it supports no diagnostic conclusion on its own.
Timing & responses
Each answer display stays up for at most 15 seconds; an unanswered trial scores 0 points. Response time is measured from the onset of the answer display, and whether you answered by keyboard (1–5), mouse, or touch is recorded per trial.
Scoring & report
Scores are recomputed on the server from your recorded trial responses: a raw total out of 60 with per-block accuracy and a full per-trial breakdown. The score is deliberately not norm-referenced — no percentile or band is shown.