Single latent factor models fitted both task success and ideal manipulation well, suggesting robust internal consistency, though Cling-subtract loaded weakly on task success¶

Claim · All claims
Evidence · 1 study (2 entries) · design r2 · q3
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? Single latent factor CFA models fitted both task success and ideal manipulation measures well, indicating robust internal consistency across the nine tasks. → Guo 2025 q3 i? The Cling-subtract task had a low factor loading (0.17) on task success, suggesting it is less effective and subject to improvement. → Guo 2025 (2)

Evidence¶

Guo 2025¶

Guo, J.; Bai, H.; Long, X.; Su, X.; Pang, W. (2025). The process of tool innovation in young children: An attempt to bridge the Gestalt and the perception-action theories. Learning and Instruction, 99, article 102182. https://doi.org/10.1016/j.learninstruc.2025.102182

q3 · i? · design · r2

CFA with WLSMV estimation on dichotomous task data from 139 children; revised measurement models allowed correlated error terms and fit significantly better than the initial models for both measures.

"Single latent factor models fitted well both measures of task success and ideal manipulation, suggesting robust internal consistency in measuring children's tool innovation performance."

Guo 2025 (2)¶

Guo, J.; Bai, H.; Long, X.; Su, X.; Pang, W. (2025). The process of tool innovation in young children: An attempt to bridge the Gestalt and the perception-action theories. Learning and Instruction, 99, article 102182. https://doi.org/10.1016/j.learninstruc.2025.102182

q3 · i? · design · r2

In the revised task-success CFA model, the Cling-subtract loading was 0.17, which the authors flag as a weakly performing task needing further improvement.

"Note, however, the factor loading of Cling-subtract was lower than the factor loadings of other tasks (0.17), indicating that this task might be less effective in measuring task success and is subject to further improvement."

Discussion¶

Related Claims¶

• Children's age, general intelligence, and divergent thinking positively predicted tool innovation performance, while executive function measures predicted only ideal manipulation — related
• Task difficulty moderated the facilitating effect of flexibility on ideal manipulation only, with the effect more salient for harder tasks — related
• The nine new tool innovation tasks showed varied difficulty levels, with task success rates ranging from 33% to 94% and ideal manipulation rates from 24% to 91% — related