Search for: All records

Spatial skills in early childhood are key predictors of mathematical achievement. Previous studies have found that training mental rotation can transfer to arithmetic skills; however, some studies have failed to replicate this transfer effect, or observed transfer effects only in certain types of arithmetic problems. Even in studies where transfer effects were observed, the underlying mechanisms of this transfer have not been explored. This study focused on the effect of short-duration (i.e., single-session) spatial training on arithmetic skills, and tested two underlying mechanisms. First, based on the spatial modeling account, short-duration spatial training may prime spatial processing, leading to a reduction in the use of counting strategies and an increase in spatially-related strategies following spatial training. Second, from a social-psychological account, short-duration spatial training may reduce children’s state anxiety, thus allowing them more cognitive resources in spatial and arithmetic tasks. We tested these mechanisms among 80 U.S. second- and third-graders using a pretest-intervention-posttest design, with 40 children in the spatial training group and 40 in an active control group. Short-duration spatial training improved children’s overall arithmetic performance; this effect did not differ by problem type (conventional, missing-term, or two-step problems). Spatial training also reduced children’s use of counting strategies. However, we did not find a significant increase in spatially-related strategies, nor did we observe a significant reduction in state anxiety. This study makes an important contribution to understanding the mechanisms underlying the transfer effects of short-duration spatial training on arithmetic skills, providing partial support for the spatial modeling account. 

The frequency of home spatial activities (e.g., puzzles and blocks) correlates with young children's spatial skills, but causal evidence is limited. We addressed this issue by comparing the effects of a parent‐led intervention aimed at increasing spatial activities to an active control targeting narrative activities (preregistered:https://osf.io/u7qrx). Parents of 80 4‐ and 5‐year‐old children were randomly assigned to either a spatial or narrative condition. Parents learned about the importance and malleability of spatial or narrative skills and engaged their children in spatial or narrative activities provided by the researchers for a month. Unexpectedly, the spatial intervention did not significantly enhance children's spatial skills or parents' motivational beliefs regarding children's spatial abilities. These findings do not support the hypothesis that spatial play causally influences children's skills. However, we note that the families in our sample had high socioeconomic status, and their children may have already benefited from rich spatial environments. 

Abstract Current biotechnologies can simultaneously measure multiple high-dimensional modalities (e.g., RNA, DNA accessibility, and protein) from the same cells. A combination of different analytical tasks (e.g., multi-modal integration and cross-modal analysis) is required to comprehensively understand such data, inferring how gene regulation drives biological diversity and functions. However, current analytical methods are designed to perform a single task, only providing a partial picture of the multi-modal data. Here, we present UnitedNet, an explainable multi-task deep neural network capable of integrating different tasks to analyze single-cell multi-modality data. Applied to various multi-modality datasets (e.g., Patch-seq, multiome ATAC + gene expression, and spatial transcriptomics), UnitedNet demonstrates similar or better accuracy in multi-modal integration and cross-modal prediction compared with state-of-the-art methods. Moreover, by dissecting the trained UnitedNet with the explainable machine learning algorithm, we can directly quantify the relationship between gene expression and other modalities with cell-type specificity. UnitedNet is a comprehensive end-to-end framework that could be broadly applicable to single-cell multi-modality biology. This framework has the potential to facilitate the discovery of cell-type-specific regulation kinetics across transcriptomics and other modalities.