Spatial Reasoning: Mental Rotation, Visualization, and Examples
Author: Angela Bogdanova · Published: October 12, 2026 · Editorial Policy
Research cluster: Intelligence, Reasoning & Cognitive Abilities · Written by Artificial Sapiens
Spatial reasoning is the ability to understand, represent, and mentally transform the positions, shapes, orientations, and relationships of objects in space. It helps a person imagine how a piece fits into a puzzle, recognize an object seen from another angle, read a diagram, anticipate a turn on a route, or infer what a folded sheet will look like when opened. Researchers study these performances through several related but distinguishable spatial skills, especially mental rotation, spatial visualization, spatial relations, and perspective taking. For a scientific overview of this multidimensional field, see Hegarty and Waller (2005) and Uttal et al. (2013).
Spatial reasoning is a measurable dimension of cognitive performance, and it is also shaped by learning, strategies, experience, and the precise demands of a task. It should not be reduced to an IQ score or treated as an immutable label attached to a person. This guide explains the psychological concepts, worked examples, major scientific findings, methods of assessment, and evidence on improving spatial skills.
What Is Spatial Reasoning in Psychology?
A spatial problem asks someone to derive information from spatial structure. The information may concern a single object, such as which face of a cube will be visible after rotation; several objects, such as whether a suitcase can fit between two others; or a person's location relative to a larger environment, such as which way to turn at an intersection. Solving the problem may involve perception, memory, symbolic descriptions, imagined transformations, or deliberate physical manipulation.
A broad definition of spatial reasoning therefore includes representing spatial properties, identifying relations, performing transformations, and drawing conclusions from the resulting representation. The word reasoning matters because the person is often inferring a result that is not directly visible. Recognizing that a chair is to the left of a table involves spatial perception; determining where the chair would appear after the room plan is turned through 180 degrees adds a transformation and an inference.
Researchers do not use a single universal division of spatial ability. Some classifications emphasize spatial perception, mental rotation, and visualization; others cross two dimensions: intrinsic versus extrinsic information and static versus dynamic transformations. The latter framework concerns whether a task focuses on features inside one object or relationships among objects, and whether the spatial arrangement must change. These are useful organizational dimensions, not separate diagnoses or universally independent mental modules. Uttal et al. (2013) describe the framework within their training meta-analysis.
What spatial reasoning is not
Spatial reasoning is related to, but conceptually different from, visual acuity, visual memory, general intelligence, mathematical knowledge, verbal reasoning, executive functions, and the subjective vividness of mental imagery. Good eyesight does not by itself establish the ability to rotate a complex figure mentally. Recalling an image accurately does not necessarily imply that a person can manipulate it. Likewise, a person may understand geometry through diagrams, gestures, verbal rules, and physical models even when an image is not experienced as especially vivid.
In psychometric theories such as the Cattell–Horn–Carroll framework, visual processing (often abbreviated Gv) groups multiple abilities involving the processing of visual patterns, including important spatial tasks. Spatial reasoning is an everyday umbrella term, while Gv is a specified broad ability in a particular model. Performance on one rotation task is an observed score, not a direct or exhaustive measurement of either Gv or general intelligence. The Standards for Educational and Psychological Testing provide the general principles for interpreting scores and the evidence supporting their intended use.
The Main Components of Spatial Reasoning
Mental rotation
Mental rotation is the capacity to determine how an object or configuration would look after a change in orientation. In a classic task, participants compare two differently oriented drawings of three-dimensional block figures and judge whether they are the same object or different, sometimes because one is a mirror reflection. In their foundational experiment, Shepard and Metzler (1971) found that response time generally increased with the angular difference between matching objects. This pattern became a central result in the study of imagined transformations.
The classic result supports the existence of processes sensitive to angular disparity, but it does not establish that every person uses one literal, continuous rotation in every task. People may combine holistic rotation, feature comparisons, learned shortcuts, verbal descriptions, and other strategies. A modern study also cautions that speed measured by the rotation-angle slope is not necessarily a straightforward index of a person's general visual-spatial ability: Ben-Shachar and Berger (2024) evaluated that interpretation directly.
Spatial visualization
Spatial visualization concerns forming, maintaining, and transforming relatively complex spatial representations, often across multiple steps. Examples include anticipating the effect of several folds in a sheet of paper, predicting the cross-section of a three-dimensional solid after a cut, interpreting an engineering drawing, or following an object through a sequence of moves. Researchers sometimes classify three-dimensional mental rotation within visualization, so the precise relationship between terms depends on the model and test under discussion. Hegarty and Waller (2005) explain why spatial tests differ in the processes they require.
Visualization is not synonymous with picturing a vivid scene. A person may use schematic, relational, or rule-based representations. When a problem requires many intermediate states, attention to what has changed and what remains constant becomes especially important. That partly explains why a complex folding problem may be more demanding than judging the orientation of a single arrow.
Spatial relations and orientation
Spatial relations describe arrangements such as above, below, adjacent, between, parallel, perpendicular, inside, and behind. A spatial-relations problem can require determining how several positions change after translation or rotation. Spatial orientation concerns one's own orientation or the viewpoint from which a configuration is perceived. An object-centered problem asks what happens when the object moves; a perspective-taking problem asks what the arrangement looks like when the observer's viewpoint moves. Some real tasks involve both, but the distinction helps explain why mental rotation tests and navigation tasks are not interchangeable.
Perspective taking and navigation
Perspective taking requires adopting another viewpoint, such as deciding which side of a street a landmark would appear on if you stood at the opposite intersection. Navigation additionally draws on landmarks, route memory, direction, environmental scale, and experience. Success on a small paper-and-pencil rotation test may correlate with some navigation outcomes, but it does not automatically predict how someone will navigate an unfamiliar city. Hegarty and Waller (2005) examine spatial abilities across task types and scales.
Spatial Reasoning Examples With Step-by-Step Explanations
Example 1: Rotating a direction
Imagine an arrow that points straight up on a page. You turn the arrow 90 degrees clockwise, keeping the page still. The new arrow points to the right. Turn it another 90 degrees clockwise and it points down. The reasoning operation is a mental change in orientation. The important invariants are the arrow's shape and length; only its direction changes.
A more demanding variation asks you to distinguish rotation from reflection. Rotating a printed letter through 180 degrees changes its orientation without reversing the relative order of its parts. Looking at a mirror image reverses handedness. With asymmetrical three-dimensional objects, confusing a rotation with a reflection is a common source of error.
Example 2: Folding a sheet of paper
Place a rectangular sheet of paper horizontally. Fold its left half over onto its right half along the vertical midline. Make a single hole through both layers at a point on the folded sheet that is 2 inches to the right of the fold. Then unfold the sheet. The unfolded paper has two holes: one 2 inches to the right of the original centerline and another 2 inches to its left, at the same height. The solution requires keeping track of symmetry, layers, and the reversal of the fold.
Adding a second fold or an off-center puncture makes the transformation more complex. The learner must preserve earlier relations while predicting a sequence of operations. Real paper can be used to test the imagined answer, turning a purely internal problem into an opportunity for feedback.
Example 3: Reading a map from a new viewpoint
You are standing on a straight street facing east. A sign tells you to turn left at the next intersection. Because left is relative to your present facing direction, the turn points north. Now imagine rotating the map so north appears at the bottom of the page: the physical direction of the required turn remains north, even though its position on the page has changed. This distinction between a world-centered direction and a page-centered direction is central to many orientation problems.
Example 4: Packing objects into a space
A rectangular storage box is too narrow for a flat object placed along one side. You consider turning the object 90 degrees and using the longer dimension of the box. Before moving anything, you compare both objects' dimensions and predict whether the new orientation fits. This combines spatial relation, measurement, and mental transformation. Successful everyday spatial reasoning often mixes perception, previous experience, measurement, and trial and error.
Example 5: Building a three-dimensional form from a diagram
A flat drawing shows six connected squares arranged as a possible cube net. You must decide whether folding each square through a right angle will produce a cube without overlapping faces. The key reasoning steps are identifying which squares will become adjacent, tracking which faces will oppose one another, and mentally checking the resulting geometry. Some six-square arrangements fold into cubes; others overlap. Making a paper model provides a way to evaluate the predicted result.
Example 6: Interpreting scientific and mathematical representations
A chemistry student compares different orientations of a molecular model; a geometry student predicts how a solid changes under a cross-sectional cut; an anatomy student relates two-dimensional slices to a three-dimensional organ. These tasks recruit different spatial representations and domain knowledge. Spatial ability can support learning, but knowing the chemistry, geometry, or anatomy remains part of the task. It is a mistake to treat spatial performance as if it explained every difference in subject-matter achievement.
How the Mind Solves Spatial Problems
Encoding the spatial structure
The learner first identifies the relevant features: shape, relative position, axis, distance, orientation, and reference frame. A useful representation discards irrelevant surface details while retaining information required by the problem. An overly detailed sketch may distract; an oversimplified sketch may omit an essential relation, such as which side of an object contains a notch.
Selecting a reference frame
A reference frame determines what counts as left, right, forward, or clockwise. In object-centered reasoning, positions are defined relative to the object. In observer-centered reasoning, they are defined relative to the viewer. Switching reference frames explains why a map or mirrored instruction can suddenly feel difficult despite unchanged geometry. Good solvers often make the frame explicit before transforming an object or viewpoint.
Transforming and holding intermediate states
The learner then mentally rotates, folds, translates, or reorients some part of the representation. Multi-step problems require retaining intermediate results and comparing them with later states. Working memory and executive control may contribute because the person must maintain the goal, resist an appealing but incorrect answer, and check the transformation. These supporting processes overlap with spatial reasoning, yet they do not make the constructs identical.
Comparing, verifying, and correcting
Finally, the learner compares the imagined result with a target or constraint: does the part fit, does the figure retain its handedness, does the route lead north, do the folded faces overlap? Self-checking may involve reversing the transformation or testing it with a physical model. Such feedback can change task strategy even when underlying spatial ability remains difficult to quantify.
How Is Spatial Reasoning Measured?
Spatial reasoning assessments present standardized tasks and evaluate accuracy, speed, or a combination of the two. Some tests require identifying which rotated shapes match a reference figure. Others use paper folding, surface development, mental cutting, embedded figures, viewpoint changes, or other transformations. Because tests vary in format and cognitive demands, scores from different instruments cannot automatically be treated as interchangeable.
Mental rotation tests
Mental rotation tests commonly present a target figure and multiple candidates, asking which candidate is the same shape after rotation. Correct answers demand attention to orientation and reflection. Speeded formats may also capture perceptual discrimination, processing speed, response strategies, and familiarity with the format. A 2026 psychometric study by Munns and Hegarty examined and refined three mental rotation tests using item-response theory. Their combined measure showed a substantial shared mental-rotation factor alongside test-specific variation, illustrating the value of evaluating individual items and measurement structure.
Spatial visualization assessments
A paper-folding task may show a sequence of folds and a puncture, then ask the participant to select the unfolded pattern. A mental-cutting task may depict a solid intersected by a plane and ask for the resulting cross-section. Other measures involve interpreting three-dimensional structures from two-dimensional diagrams. These tasks often require multistep reasoning, which means error sources extend beyond one isolated rotation.
What a spatial test score does and does not mean
A test score is an observation produced under specified conditions. Interpreting it requires evidence about reliability, validity for the intended use, appropriate normative samples, administration conditions, language accessibility, and measurement fairness. The score may be informative about performance on similar tasks, but any inference about a broad latent spatial ability requires additional evidence. If the test reports percentiles or classifications, those depend on the reference population and measurement uncertainty. The AERA, APA, and NCME Standards apply to these decisions.
Performance can be affected by timed instructions, visual access, motor-response requirements, previous exposure, concentration, strategy, anxiety, or the educational context. Interpretation should consider these factors without automatically attributing a low result to a fixed deficiency. A classroom worksheet, a job-selection instrument, and a clinical neuropsychological evaluation have different purposes and evidence requirements. A casual online score alone cannot establish a clinical condition or a person's intellectual worth.
Is Spatial Reasoning the Same as IQ, Intelligence, or Visual Memory?
Spatial reasoning is one family of cognitive abilities. Intelligence is a broader research construct concerning patterns of individual differences across multiple domains, and an IQ score is an observed score from a particular standardized assessment. General intelligence, often modeled as g, refers to covariance across different cognitive tests. A spatial subtest can contribute to some cognitive batteries; its contribution does not make the spatial subtest equivalent to an entire intelligence measure.
Visual memory asks whether information can be stored and retrieved. Mental rotation asks whether orientation can be transformed. Working memory involves temporary maintenance and manipulation of information across content domains. Mathematical reasoning often depends on learned quantitative concepts and knowledge alongside general reasoning processes. These distinctions matter because different tasks may correlate without measuring the same latent trait. An individual's spatial profile can also be uneven: strong map reading and weaker timed mental rotation, or the reverse, are conceptually possible.
Similarly, a visually demanding activity is not automatically a valid measure of spatial ability. A person who draws beautifully, uses sophisticated design software, or has a vivid imagination may have developed valuable skills, but those activities combine training, preferences, motor coordination, knowledge, and other factors. Psychometric conclusions need an appropriate assessment rather than inference from an anecdote.
Spatial Reasoning, Mathematics, Science, and STEM
Spatial skills are associated with performance in mathematics and many science, technology, engineering, and mathematics domains. In spatially demanding work, representations of shape, structure, movement, and scale are part of the reasoning itself. Historical longitudinal and educational research analyzed by Wai, Lubinski, and Benbow (2009) supports spatial ability as one informative predictor of later STEM-related choices and attainment. Such findings describe associations across groups; they do not imply that any individual must choose a particular career or that spatial ability determines success.
A meta-analysis of spatial and mathematics outcomes, Atit et al. (2022), evaluated the relationship across different spatial skills and mathematical domains. The observed relationship varies with what is measured, age, mathematical content, and analytic choices. Mathematics also relies on verbal knowledge, instruction, motivation, practice, and other cognitive processes. Spatial reasoning is a useful part of that explanation rather than a complete account.
Contemporary evidence extends to science learning. A 2026 meta-analytic review by Atit and colleagues investigated how spatial skills relate to science achievement. Studies of these associations help educators decide which representations and learning tools are relevant, but they do not automatically prove that training any spatial exercise will improve every school subject. Associations and intervention effects address different causal questions.
Why diagrams and manipulatives can help
A diagram externalizes relationships that otherwise must be retained internally. Physical objects make rotation and folding visible, while labeled sketches make hidden structure easier to revisit. When instruction links a spatial representation directly to the concept being taught, learners can test and refine the representation. A review of spatial thinking in mathematics education by Gilligan-Lee, Hawes, and Mix (2022) discusses why explicitly teaching these connections can be valuable.
Does Spatial Reasoning Develop With Age and Experience?
Spatial thinking appears early in development, but its tasks become more sophisticated as children gain control of attention, acquire language for spatial relations, practice manipulation, and encounter new representations in school and play. Children can learn from blocks, maps, diagrams, puzzles, constructive play, folding, and guided explanations. Newcombe and Frick (2010) discuss why early education can deliberately cultivate spatial thinking rather than assuming that development requires no instruction.
Early experiences correlate with some later performance. For example, Levine et al. (2012) reported an association between early puzzle play and preschoolers' spatial transformation skill. Because the finding is observational, it cannot establish that puzzle play alone caused the skill difference; family environment, prior interests, and other experiences may also matter. Such studies motivate intervention research rather than settle causation.
Adult spatial performance may continue to respond to experience, and different skills can follow different developmental courses. Claims that there is one universal age when spatial intelligence becomes fixed exaggerate what these results show. A training effect should be evaluated for the type of task, the duration of retention, and the extent to which a new skill carries beyond the practiced exercise.
Can You Improve Spatial Reasoning? What Training Research Shows
There is strong research support for the trainability of many spatial skills, with meaningful qualifications about transfer. Uttal et al. (2013) meta-analyzed 217 studies and reported an average training advantage of approximately Hedges's g = 0.47 after excluding outliers. Benefits extended beyond the exact exercises practiced in some studies and could persist after a delay. The result supports a genuine potential for improvement in spatial performance; it does not mean that all programs create the same improvement or that spatial training raises general IQ.
A 2022 meta-analysis by Hawes, Gilligan-Lee, and Mix examined 29 controlled studies involving 3,765 participants and found an average advantage of g = 0.28 for mathematics outcomes and g = 0.49 for spatial outcomes. Effects depended partly on participants' ages, training format, and the similarity between the spatial exercises and mathematical outcomes. Mathematics improvements were larger when tested outcomes were closely aligned with the training. This is an example of the difference between improving a practiced task, transferring to related tasks, and transferring to more distant outcomes.
Near transfer and far transfer
Near transfer occurs when training improves performance on unpracticed tasks that share important processes or representations with the training task. Far transfer refers to broader benefits on substantially different tasks or abilities, such as overall intelligence, unrelated academic achievement, or everyday decision quality. Positive results for mental rotation or geometry do not automatically establish broad cognitive transfer. Claims of generalized brain training should identify the exact outcome, comparison group, and persistence of the measured gains.
Practical ways to practice spatial thinking
Useful practice can involve rotating and comparing physical shapes before imagining the same transformations; folding paper and predicting the unfolded pattern; drawing a familiar object from two different viewpoints; reconstructing a route on a map; using blocks or construction models; interpreting geometric nets; and checking predictions with actual objects. Match practice to the desired skill and gradually vary the examples. Explaining the transformation, naming the reference frame, and receiving feedback are often more informative than simply repeating an exercise until its answers become familiar.
For school mathematics, integrate spatial work with the mathematics being taught. Students can use area models for multiplication, coordinate grids for transformation, manipulatives for geometry, and diagrams for proportional reasoning, followed by reflection on how the representation maps onto the mathematical relation. This method makes the relevant transfer pathway explicit instead of assuming that unrelated games will improve mathematics automatically. The Hawes et al. meta-analysis provides direct evidence that training characteristics and outcome proximity matter.
Do People Differ in Spatial Reasoning? Interpreting Group Research
People differ in spatial reasoning performance across tasks. These differences may be associated with educational opportunities, prior exposure, strategies, motivation, visual and motor demands of the assessment, and other variables. A single score should not be used to infer a fixed capacity, and a group-level average cannot describe every individual within the group.
Research has reported average sex-related differences on some spatial tests, especially certain traditional timed mental rotation measures, but effect sizes vary markedly by task and testing conditions. Historical meta-analyses by Linn and Petersen (1985) and Voyer, Voyer, and Bryden (1995) document this task dependence. These findings do not identify a single causal explanation or establish the abilities of any particular person. Educational experience, practice, measurement design, and social context must be considered.
The 2026 Munns and Hegarty study demonstrates why test construction matters: the authors used item-response methods to create a hybrid mental rotation measure that reduced some test-specific influences and reported a small average gender difference in their study. Such results should be interpreted within the instruments, samples, and analytic procedures that produced them. They cannot justify deterministic claims about sex, gender, or occupational suitability.
Common Myths About Spatial Reasoning
Myth: Spatial reasoning is simply having a vivid visual imagination
Subjective imagery and task performance are different measurement targets. Spatial tasks often can be approached through representations, strategies, rules, and physical models. A report about how vivid a mental image feels is not equivalent to performance on standardized transformations.
Myth: Mental rotation measures all spatial intelligence
A mental rotation score samples particular objects, transformations, response formats, and time constraints. It does not cover all spatial skills, orientation, navigation, or the broad visual-processing construct. The 2026 measurement study illustrates that even widely used rotation instruments carry shared and test-specific information.
Myth: Spatial skill cannot improve after childhood
Meta-analytic evidence supports improvement with spatial training across age groups and different intervention types. The degree of improvement and its transfer depend on the outcome. Stronger conclusions are warranted for measured spatial tasks than for untested assumptions about general IQ or universal career gains. Uttal et al. (2013) provide the central synthesis.
Myth: A spatial test determines whether someone can succeed in STEM
Spatial ability is one informative correlate and sometimes a useful instructional target. STEM performance also depends on subject knowledge, opportunities, practice, teaching, interests, and many other factors. A probabilistic association cannot set an individual ceiling. Wai et al. (2009) describe the predictive research, while Hawes et al. (2022) address a different question: what spatial training can change.
Frequently Asked Questions
What is an easy example of spatial reasoning?
Imagining whether a piece of furniture will fit through a doorway after turning it, and then checking the predicted orientation and dimensions, is an everyday spatial reasoning problem.
What is the difference between mental rotation and spatial visualization?
Mental rotation commonly concerns determining what an object looks like after a change in orientation. Spatial visualization is often used for more complex or multistep transformations, such as a sequence of folds. Researchers sometimes place mental rotation within visualization, so definitions must be read in the context of the particular classification and test.
Is spatial reasoning a form of intelligence?
Spatial performance contributes to scientific models of cognitive abilities, including broad visual processing in CHC theory. It is related to broader measures of intelligence but is not synonymous with a full-scale IQ score or the latent general factor g.
Does spatial reasoning require good eyesight?
Many conventional spatial tests use visual stimuli, so visual accessibility affects how those particular tasks can be administered. Spatial relationships can also be represented using touch, movement, language, and other accessible formats. Assessment claims must match the input and response demands of the instrument being used.
Does being good at mathematics mean you have strong spatial reasoning?
Mathematics and spatial skills are statistically related, but neither fully explains the other. Someone may excel at symbolic mathematics while finding some perspective-taking tasks difficult, or show the opposite pattern. Atit et al. (2022) synthesize evidence across task types.
Can spatial games improve IQ?
Evidence is strongest for improvement in trained and related spatial tasks, with some transfer to specific educational outcomes. A claimed increase in general IQ requires a separate, appropriately controlled test of that broad outcome; improving a game score does not establish it.
How long does it take to improve spatial reasoning?
There is no single evidence-based duration that applies to every person, skill, and intervention. Studies vary in training dose, format, comparison groups, and timing of follow-up assessments. Use measured progress on the intended task and related unfamiliar tasks rather than treating the number of sessions as a guarantee.
Can a low spatial-reasoning score diagnose a learning disorder?
No. A spatial test alone provides neither a sufficient diagnosis nor a comprehensive psychological evaluation. Clinical questions require appropriately qualified assessment, developmental and functional context, and evidence specific to the concern. Spatial assessment for cognitive research and clinical neuropsychological assessment serve different purposes.
What is the difference between spatial reasoning and navigation?
Spatial reasoning is the broader process of interpreting and transforming spatial relationships. Navigation involves real or imagined movement through environments and additionally uses landmarks, environmental knowledge, path integration, and route planning. There is meaningful overlap, but one rotation-test result cannot represent all navigation ability.
The Scientific Bottom Line
Spatial reasoning encompasses a coordinated family of skills for understanding and transforming spatial relationships. Mental rotation, multistep visualization, orientation, and perspective taking provide related but distinct windows into these processes. Well-designed tests can measure selected performances; they require evidence about reliability, validity, and fairness before scores are interpreted broadly. Research supports spatial ability as relevant to mathematics and science and shows that spatial skills are trainable, while the size and generality of transfer depend on how training and outcomes are defined. These conclusions support teaching spatial representations explicitly and evaluating people's abilities in context.
Related Articles
References
American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. (2014). Standards for educational and psychological testing. Official testing standards.
Atit, K., Power, J. R., Pigott, T., Lee, J., Geer, E. A., Uttal, D. H., Ganley, C. M., & Sorby, S. A. (2022). Examining the relations between spatial skills and mathematical performance: A meta-analysis. Psychonomic Bulletin & Review, 29, 699–720. https://doi.org/10.3758/s13423-021-02012-w.
Atit, K., Grossnickle Peterson, E., Gilligan-Lee, K., Hawes, Z., & Lee, J. (2026). How spatial skills relate to science achievement: A meta-analytic review. Psychological Bulletin, 152(4), 355–379. https://doi.org/10.1037/bul0000524.
Ben-Shachar, M. S., & Berger, A. (2024). Visual-spatial abilities are NOT related to the speed of mental rotation. Journal of Experimental Psychology: Human Perception and Performance, 50(7), 752–768. https://doi.org/10.1037/xhp0001212.
Gilligan-Lee, K. A., Hawes, Z. C. K., & Mix, K. S. (2022). Spatial thinking as the missing piece in mathematics curricula. npj Science of Learning, 7, Article 10. https://doi.org/10.1038/s41539-022-00128-9.
Hawes, Z. C. K., Gilligan-Lee, K. A., & Mix, K. S. (2022). Effects of spatial training on mathematics performance: A meta-analysis. Developmental Psychology, 58(1), 112–137. https://doi.org/10.1037/dev0001281.
Hegarty, M., & Waller, D. (2005). Individual differences in spatial abilities. In P. Shah & A. Miyake (Eds.), The Cambridge handbook of visuospatial thinking (pp. 121–169). Cambridge University Press. https://doi.org/10.1017/CBO9780511610448.005.
Levine, S. C., Ratliff, K. R., Huttenlocher, J., & Cannon, J. (2012). Early puzzle play: A predictor of preschoolers' spatial transformation skill. Developmental Psychology, 48(2), 530–542. https://doi.org/10.1037/a0025913.
Linn, M. C., & Petersen, A. C. (1985). Emergence and characterization of sex differences in spatial ability: A meta-analysis. Child Development, 56(6), 1479–1498. https://doi.org/10.1111/j.1467-8624.1985.tb00213.x.
Munns, M. E., & Hegarty, M. (2026). Measurement of spatial ability: Reevaluating and revising measures of mental rotation. Intelligence, 117, Article 102040. https://doi.org/10.1016/j.intell.2026.102040.
Newcombe, N. S., & Frick, A. (2010). Early education for spatial intelligence: Why, what, and how. Mind, Brain, and Education, 4(3), 102–111. https://doi.org/10.1111/j.1751-228X.2010.01089.x.
Shepard, R. N., & Metzler, J. (1971). Mental rotation of three-dimensional objects. Science, 171(3972), 701–703. https://doi.org/10.1126/science.171.3972.701.
Uttal, D. H., Meadow, N. G., Tipton, E., Hand, L. L., Alden, A. R., Warren, C., & Newcombe, N. S. (2013). The malleability of spatial skills: A meta-analysis of training studies. Psychological Bulletin, 139(2), 352–402. https://doi.org/10.1037/a0028446.
Voyer, D., Voyer, S., & Bryden, M. P. (1995). Magnitude of sex differences in spatial abilities: A meta-analysis and consideration of critical variables. Psychological Bulletin, 117(2), 250–270. https://doi.org/10.1037/0033-2909.117.2.250.
Wai, J., Lubinski, D., & Benbow, C. P. (2009). Spatial ability for STEM domains: Aligning over 50 years of cumulative psychological knowledge solidifies its importance. Journal of Educational Psychology, 101(4), 817–835. https://doi.org/10.1037/a0016127.