Reasoning4–6 minBurden: MediumEMA: Low

Cube Comparison

A spatial visualisation task: two drawn cubes with a symbol on each face; can they be the same cube, turned? Every pair is generated from a seed.

Spatial visualisationMental rotationSpatial ability
Category
Reasoning
Typical duration
4–6 min
Participant burden
Medium
EMA suitability
Low

Try this task

Configure parameters and run an interactive preview exactly as participants will experience it. No data is recorded.

Configure preview

Adjust parameters below, then start the preview on the right.

Include practice trials

Shown with feedback before the main task

Task parameters

Shared evenly across the levels, easiest first. Default 21, as one part of the Cube Comparisons test of the ETS Kit of Factor-Referenced Cognitive Tests (Ekstrom, French, Harman & Dermen, 1976; two parts of 21 items, as described by He et al., 2022, and Stieff et al., 2020).

With feedback, after two worked examples; two per level in turn, one same and one different. Default 6, as in Just & Carpenter (1985).

Levels

The turn that brings the left cube into the position of the right one, in quarter turns about the cube's axes, as in the analysis of Cube Comparisons items by Just & Carpenter (1985), where response times and errors rose with it. Level 1: a quarter turn (two faces shared). Level 2: two quarter turns: a half turn about one axis or quarter turns about two axes (one face shared), or a twist about the corner both drawings show (three faces shared), in turn. Level 3: three quarter turns, a flip about an edge (two faces shared).

Default 0.5: as many same as different pairs, as in Just & Carpenter (1985).

Empty: each participant gets a new form, and the seed it was drawn from is stored with the result. The same seed gives every participant the same pairs.

Default none (our choice: a power measure with RTs). A pair not answered in time counts as an omission.

Default none. The Kit gives 3 min for each part of 21 items: 180 s with the default 21 pairs gives its pace. The clock starts with the first live pair and is shown; the pair on screen at the end is kept but not scored, and pairs never shown count as not reached. Practice is not timed.

Two cubes per pair, three faces visible on each, a symbol with a direction on every face (no symbol twice on one cube); the participant taps Same or Different. A different pair is a same pair with one change to the right cube, as Just & Carpenter (1985) built them: a shared symbol turned, two symbols swapped, or a shared symbol replaced by the left cube's other symbol. Each change keeps the number of symbols the two cubes share, so neither that number nor a symbol missing from one cube gives the answer away. About 5 min with these settings. Outputs: correct, errors, answered, correct minus errors (the Kit's score), not reached (with a time limit for all pairs), accuracy for same and different pairs, per change and per level, median RT of correct answers; each trial stores both cubes, the turn between them and the change.

4–6 minBurden: MediumEMA: Low

This is a researcher preview. No participant data is recorded.

Simulated participant view

9:41

Cube Comparison

A spatial visualisation task: two drawn cubes with a symbol on each face; can they be the same cube, turned? Every pair is generated from a seed.

No data is recorded

Participant experience on smartphone

Two cubes side by side in isometric view, each showing three faces with a filled symbol on each, and Same and Different buttons below, answered at touch-down. Symbols are dark shapes on white or light grey faces, so nothing depends on colour.

When to use

For a spatial visualisation measure in the format of the Cube Comparisons test of the ETS Kit of Factor-Referenced Cognitive Tests, with many parallel forms: pairs are generated, so sessions can use new pairs (no seed) or one fixed form (a set seed). Suited to baseline and occasional sessions, and to studies of the cube items used in aptitude and selection tests.

When not to use

Not the Kit's test itself and not normed: the pairs follow its format and the item structure analysed by Just & Carpenter (1985), but have no norms. Not a quick EMA probe. For folding a net into a cube, use Cube Folding; for 2-D rotation, Mental Rotation.

How to use in a study

The defaults give 21 pairs, 7 per level, levels 1 to 3, easiest first, half of them the same cube, without time limits, after two worked examples and 6 practice pairs with feedback. For the Kit's time pressure, set a time limit of 180 s for all pairs (3 min per part of 21 items): the score is then correct minus errors, and pairs not reached are counted. Leave the seed empty to give each participant and each session new pairs; set a seed to give everyone the same form.

Researcher-configurable parameters

  • Number of pairs: default 21, as one part of the Kit's test (two parts of 21 items)
  • Levels: 1 one quarter turn between the cubes, 2 two quarter turns, 3 three quarter turns (Just & Carpenter, 1985)
  • Proportion of same pairs: default 0.5 (Just & Carpenter, 1985)
  • Time limit for all pairs: default none; 180 s with 21 pairs gives the Kit's pace (3 min per part)
  • Time limit per pair: default none (our choice)
  • Seed: empty (new pairs per run; the seed is stored with the result) or a number (the same pairs for everyone)
  • Practice pairs with feedback: default 6 (Just & Carpenter, 1985), two per level, one same and one different

Outputs collected

  • Correct answers, errors, pairs answered, and correct minus errors (the Kit's score)
  • Pairs not reached, with a time limit for all pairs
  • Accuracy for same pairs and for different pairs (chance 50 %)
  • Accuracy per change in different pairs (a symbol turned, two symbols swapped, a symbol replaced) and per level
  • Median RT of correct answers
  • Every trial: level, problem type (the turn between the cubes and the faces they share), both cubes as drawn, the cube they come from and its two rotations, the change in a different pair, the response, touch-down RT, time-outs, and the pair cut off at time-up
  • The seed, from which every pair of the run can be rebuilt

Interpretation notes

Each cube has six different symbols, one per face, all with a direction (like an arrow), so every symbol the two drawings share shows which way the cube was turned. The cubes and symbols are drawn, not copied from the Kit, and symbols replace the Kit's letters so that no script or mirror-image letter is involved. Levels follow Just & Carpenter (1985), who analysed the test's same pairs by the number of quarter turns about the cube's axes that align the two cubes and the number of faces they share; response times and errors rose with the turns, and low-spatial participants made most errors with three shared faces after two quarter turns and with three quarter turns. Level 1 is one quarter turn (two shared faces), level 2 two quarter turns (one or three shared faces, three problem types in turn), level 3 three quarter turns (two shared faces). Pairs with no shared face are not used, because any two such drawings can show one cube. A different pair is a same pair with one change to the right cube, as Just & Carpenter built theirs: a shared symbol turned, two symbols swapped, or a shared symbol replaced by the left cube's other symbol, so a symbol is missing from one cube and another sits where it cannot be. Every change keeps the number of shared symbols, so same and different pairs of a level share the same problem types, the same number of shared symbols and the same places for the symbols not shared, and no symbol is ever drawn unchanged in the same place: counting shared symbols or noticing a missing one is right at chance. With one shared face only a turned symbol can make the cubes different, so those pairs test direction only. A pair is different only when none of the 24 rotations of the left cube gives the right one, checked for every pair. Accuracy for same, different and each change has a chance level of 50 %. RTs are timed from the drawn pair to the touch-down, but this timing has not been validated on real devices. The pairs have no norms; compare groups or sessions within a study.

Scientific evidence

  • The Kit's Cube Comparisons test shows two drawings of a cube with a letter, number or symbol on each face, no letter twice on one cube; participants mark whether they can be the same cube, in two parts of 21 items with 3 min each, scored as right minus wrong (He et al., 2022; Stieff et al., 2020; Timm et al., 2024).
  • In an eye-tracking analysis of Cube Comparisons items, response times rose with the number of quarter turns needed to align the cubes; low-spatial participants were slower and made more errors, most of them where a letter had to pass a hidden face on its way (two quarter turns with three shared faces, and three quarter turns; Just & Carpenter, 1985).
  • Just & Carpenter (1985) built their different pairs from same pairs by changing the location or the direction of a matching letter or by exchanging two letters; the changes here follow that.
  • The test is a common marker of spatial visualisation alongside Paper Folding (He et al., 2022; Stieff et al., 2020).

Timing and session conditions

Each run of this task records how its times were measured and the conditions of the session, such as screen brightness, battery saving and interruptions. Runs with possible problems are flagged on the Quality tab and in the export.

How SMAAT times cognitive tasks

Read more

Related tasks

Add Cube Comparison to your ESM study on SMAAT — free to get started.