幼兒性別如何影響親子空間對話？空間語言頻率與廣度的探討

How Does Child Gender Influence Parent-Child Spatial Discourse? Exploring the Frequency and Breadth of Spatial Language

親子間的空間對話能引導幼兒關注環境中的關鍵特徵與標記空間概念，進而提升其空間推理與記憶能力。然而，某些空間技能自幼兒期起即呈現性別差異，故探討親子的空間概念討論，及這些討論內容如何受到幼兒性別影響，對於提升教育實踐的品質與公平性至關重要。本研究旨在分析親子對話中空間語言的表達內涵，包括頻率（量）與廣度（質），並探討幼兒性別對親子空間語言的影響。研究收集52對學前幼兒及其家長進行空間遊戲的語料，並依據空間語言類型架構進行編碼與分析。結果顯示，家長對男童的空間語言使用更為頻繁且多元，男童的空間遊戲時間與空間語言廣度顯著高於女童。階層迴歸分析進一步發現，幼兒的空間語言頻率與廣度皆受到家長空間語言輸入的影響，但遊戲時間對語言廣度影響更為顯著。此外，家長的空間語言輸入對女童的空間語言頻率與廣度均有顯著且更佳的促進效果。研究結果不僅有助於理解親子空間對話的特性與影響因素，也為家長支持不同性別子女的空間能力發展提供實證依據。

Early spatial ability is a critical component of intelligence assessments and is closely associated with later mathematics and reading performance after school entry (Verdine et al., 2017; Wang et al., 2021). This ability serves as a core foundation for subsequent STEM learning (Hodgkiss et al., 2018; Wai et al., 2009; Wai et al., 2010). However, specific spatial skills (e.g., mental rotation) indicate a male advantage as early as preschool-aged children (Johnson et al., 2022; Ralph et al., 2021)—a difference that may be related to spatial language use patterns.
Spatial language is a key factor influencing spatial cognition and ability (Pruden et al., 2011; Wu et al., 2022). Research suggests that parent–child spatial discourse directs children’s attention to critical features of their environment and labels spatial concepts, improving spatial reasoning and memory (Ankowski et al., 2012; Casasola et al., 2017; Miller et al., 2016). However, parents’spatial language typically varies by child gender: when interacting with boys, parents tend to use spatial language more frequently and in a greater variety of types (Pruden & Levine, 2017; Umansky & Callanan, 2024; Verdine et al., 2019). Whether such differences further influence children’s spatial language development remains underexplored. Thus, examining parent–child discourse concerning spatial concepts and whether the relationship is moderated by child gender is critical for understanding gender disparities in educational processes and for promoting educational equity.
The present study examined the features of spatial language in parent–child discourse, focusing on frequency (quantity), breadth (quality), and the role of child gender. It comprises a secondary data analysis collected initially on a project of parent–child mathematical language and children’s mathematical abilities (Huang, 2025). The source study compiled a corpus of interactions from parent–child dyads who participated in three mathematical plays and assessments of children’s mathematical ability and rapid automatized naming (RAN, as an indicator of general cognitive ability). The current analysis focused exclusively on the corpus from one play involving spatial concepts.
The sample encompasses 52 parent–child dyads. The child mean age was 62.75 months (range = 50–75 months, SD = 7.51), of which 23 were girls (44.23%). Mothers were the accompanying parent in 43 cases (82.69%) and fathers in nine cases (17.31%). Socioeconomic status was indexed by the highest parental education of either parent, categorized as less than college (7.69%), college graduate (44.23%), and postgraduate (48.08%).
The corpora were transcribed verbatim and coded at the utterance level. Coding was based on six categories of spatial language (spatial features and properties, shapes, dimensions, transformations and orientations, locations and directions, and spatial deictics; Bryant et al., 2024; Cannon et al., 2007). For each speaker, frequency (counts per category and total), breadth (number of categories covered, range = 0–6), and total counts of utterance (spatial and non-spatial) were calculated. Since the durations of the play varied, frequencies and total utterances were standardized per five minutes, while the breadth was not transformed.
The study first conducted descriptive statistics and examined the association of spatial language between parents and children. The average duration of spatial play was 568.60 seconds (SD = 258.72, range = 139–1082). Boys played significantly longer than girls (t(50) = -2.12, p = .039, d = .590). The parent spatial language frequency averaged 14.76 utterances (range = 0.73–44.77, SD = 8.59), representing 45.10% of all utterances, with an average breadth of 4.77 categories (range = 1–6, SD = 1.15). The child spatial language frequency averaged 6.29 utterances (range = 0–17.65, SD = 3.99), representing 21.06% of all utterances, with an average breadth of 3.71 categories (range = 0–6, SD = 1.56). Parents’and children’s spatial language were positively correlated in frequency (r(52) = .50, p < .001), breadth (r(52) = .43, p = .001) and total utterances (r(52) = .92, p < .001), indicating a strong alignment in parent–child language input and response.
Subsequent analyses tested potential covariates (child age by month, RAN, play duration, and total utterances of parents and children) and compared differences by parent and child gender and socioeconomic status. The parent spatial language frequency was negatively correlated with child age (r(52) = -.33, p = .018) and differed significantly by child gender (t(50) = -1.75, p = .087, d = -.488), with higher frequency when interacting with boys (M = 16.57, SD = 9.16) than with girls (M = 12.46, SD = 7.38). The parent spatial language breadth was also negatively associated with child age (r(52) = -.24, p = .082) but positively related to the duration of play (r(52) =.53, p < .001). The breadth also differed by child gender (t(55) = -2.47, p = .017, d = -.690), with a greater variety of spatial terms used with boys (M = 5.10, SD = 0.94) than with girls (M = 4.35, SD = 1.27).
Children’s spatial language frequency was positively correlated with both parents’(r(52) = .28, p = .044) and children’s total utterances (r(52) = .31, p = .025). Children’s spatial language breadth was negatively correlated with RAN (r(52) = -.27, p = .056) and positively correlated with play duration (r(52) = .60, p < .001). The breadth differed marginally by child gender (t(50) = -1.70, p = .095, d = -.476), with boys demonstrating greater breadth (M = 4.03, SD = 1.32) than girls (M = 3.30, SD = 1.55).
Finally, hierarchical multiple linear regressions were conducted with child spatial language frequency and breadth as the dependent variables. In Step 1, child age, RAN, play duration, and both parent and child total utterances were added as control variables. The results indicate that these control variables significantly predicted child spatial language breadth (43.73%, F(5, 46) = 7.15, p < .001), but not child spatial language frequency (14.76%, F(5, 46) = 1.59, p = .181).
In Step 2, parent spatial language frequency (X1) and child gender (W) were added to the frequency model. Parent spatial language breadth (X2) and child gender (W) were added to the breadth model. In the child spatial language frequency model, adding the two predictors (X1 and W) explained 22.63% of the variance (p = .001). The influence of parent spatial language frequency (X1) was significant (β= .55, t = 3.97, p < .001), with the overall model significant (F(7, 44) = 3.75, p = .003). In the child spatial language breadth model, the two predictors (X2 and W) did not make a significant contribution to the model, explaining only 4.72% (p = .146) of the variance (F(7, 44) = 5.91, p < .001).
In Step 3, the interaction term of the predictor variables was added to both models. The interaction (X1×W) was a significant predictor of child spatial language frequency (β= -.46, t = -2.23, p = .031; F(8, 43) = 4.20, p < .001). It explained an additional 6.50% (p = .031) of the variance, indicating a moderating role of child gender. In contrast, the interaction (X2×W) was not a significant predictor of child spatial language breadth (β= -.10, t = -0.60, p = .553; F(8, 43) = 5.14, p < .001) and increased variance explained by only 0.42% (p = .553).
The findings indicate that spatial play effectively promotes spatial discourse between parents and children, with approximately 45% of parent and 21% of child utterances involving spatial concepts. However, the proportion of child spatial language was lower. This is potentially because parents tended to use declarative or closed-ended questions, constraining children’s extended expression. Moreover, spatial language varied considerably between dyads. This finding suggests the value of providing parents with targeted guidance to adjust language use intentionally, encouraging children to produce spatial language while strengthening their conceptual connections.
Comparing parent and child patterns revealed that parents’spatial language was shaped by characteristics of the interaction partner, particularly child gender and age. The higher frequency and breadth of spatial language directed at boys may reflect parents’gender stereotypes and social expectations. Parents may perceive spatial activities as more suitable for boys or assume that boys possess stronger spatial abilities (Endendijk et al., 2018), thus using more spatial language with their male children.
In contrast, children’s spatial language use was more strongly associated with their general cognitive ability, total number of utterances, and play duration. This finding highlights the importance of both contextual and individual factors. Duration was particularly influential for breadth, while the frequency was more strongly correlated with parental input. Particularly, parents’spatial language showed a stronger facilitative effect on girls’spatial language. This indicates that appropriate guidance can effectively enhance girls’participation in spatial conversation.
This study contributes to understanding the characteristics and influencing factors of parent–child spatial language interactions. It provides empirical evidence to inform parental practices that support the development of spatial abilities in children of both genders.