Evidence for a unitary structure of spatial cognition beyond general intelligence

Spatial skills are crucial for navigating the world and are associated with positive developmental outcomes and success in STEM fields. Historically, spatial ability research has focused on object-based skills (mental rotation, visualization), while large-scale spatial orientation (navigation) has received less attention. Difficulties in reliably measuring navigation in real-world settings have hindered research. Recent advances in virtual reality (VR) offer a solution, allowing for standardized and controlled assessment. Existing research presents contrasting views on the relationship between object-based and spatial orientation abilities, with some suggesting distinct abilities and others proposing a unitary structure. This study aims to address this gap by investigating the structure of spatial ability across a comprehensive range of object-based and spatial orientation tests, using a large genetically informative sample of twins, to determine the heritability and the role of general cognitive ability (g).

The literature presents conflicting views on the nature of spatial ability. Some research suggests a distinction between object-based spatial skills (mental manipulation of objects) and large-scale spatial orientation (navigation), citing differences in cognitive processes and brain regions involved. Other studies argue for a largely unitary structure, supported by observed correlations between object-based and spatial orientation abilities. The role of general cognitive ability (g) in this relationship also remains unclear. While some studies demonstrate the coherence of spatial abilities independently of g, neuropsychological evidence suggests a broader impact of navigation-related brain damage. The lack of consensus is attributed to limited studies using diverse test batteries and a lack of genetically informative designs.

The study utilized data from the Twins Early Development Study (TEDS), focusing on 2660 twins aged 19-22. Two online gamified batteries were used. The first, "Spatial Spy," assessed six aspects of spatial orientation in a virtual environment: Map Reading, Route Memory, Navigation Directions, Navigation Landmarks, Large-scale Scanning, and Large-scale Perspective-taking. The second battery, "The King’s Challenge," assessed ten object-based spatial skills: Mazes, 2D Drawing, Elithorn Mazes, Pattern Assembly, Mechanical Reasoning, Paper Folding, 3D Drawing, Mental Rotation, Perspective-taking, and Cross-sections. General cognitive ability (g) was measured using a composite score from multiple tests administered at ages 7-16. Phenotypic correlations and confirmatory factor analyses were performed to examine the structure of spatial abilities. Univariate and multivariate twin modeling was used to estimate heritability and the influence of genetic and environmental factors. A Cholesky decomposition examined the role of g in the observed spatial ability structure.

The six spatial orientation tests in "Spatial Spy" clustered into a single 'Navigation' factor (64% heritable). Analysis across all 16 spatial tests revealed three substantially correlated factors: Navigation, Object Manipulation, and Visualization. These loaded strongly onto a higher-order general Spatial Ability factor (84% heritable). A large portion (45%) of this heritability was independent of g. This indicates a significant common genetic influence on spatial abilities beyond their correlation with general intelligence. Shared environmental effects were negligible; nonshared environmental variance was largely test-specific. Males outperformed females on all tests, with effect sizes ranging from small to moderate. Sex limitation modeling revealed significant quantitative sex differences for navigation ability and some subtests, but the differences were small-moderate. The three-factor hierarchical model provided a better fit than simpler models, supporting the complex yet interconnected nature of spatial abilities.

The findings strongly support a unitary structure of spatial cognition, both phenotypically and genetically, that extends beyond the influence of general cognitive ability (g). The high heritability of the general Spatial Ability factor, independent of g, suggests a shared genetic network underlying diverse spatial skills. This contradicts some theories positing distinct object-based and spatial orientation abilities. The study highlights the importance of considering the interconnectedness of different aspects of spatial cognition. The observation of three factors within a broader spatial ability emphasizes the complexity of spatial thinking, while the high correlation between them underscores the underlying unity. The discrepancy between the unitary nature of spatial abilities and individual perceptions of these abilities warrants further research.

This study provides compelling evidence for a unitary structure of spatial cognition, supported by strong phenotypic and genetic correlations across diverse spatial tasks. The significant heritability of this ability, independent of general intelligence, points to a common genetic architecture. Future research should investigate the specific genes and biological mechanisms underlying this unitary structure and explore the implications for educational interventions aimed at improving spatial skills.

The use of a virtual environment for assessing spatial orientation might not fully capture real-world navigation abilities. While the VR environment was designed to be realistic, differences between simulated and real-world navigation could exist. The study's reliance on self-report data for some aspects of spatial anxiety could introduce subjective biases. The sample is limited to UK twins, which may limit the generalizability of findings to other populations.