Review Reports - The Relationship Between Social Adaptation and Parenting Styles in Left-Behind and Non-Left-Behind Children: A Network Analysis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript investigates how left-behind children (LBC) in rural China adapt socially, drawing on a network analysis framework. The study pursues two main goals: identifying the core dimensions of social adaptation and comparing network structures between LBC and their non-left-behind peers (NLBC), while also examining how parenting styles relate to social adaptation outcomes. The sample is substantial (N = 2,452), and the analyses make use of relatively sophisticated statistical tools, including EBICglasso, centrality indices, and bridge expected influence. The manuscript's most notable contributions are its methodological originality, its theoretically grounded approach, and its effort to model parenting styles and adaptive outcomes within a single framework.

The topic is relevant and timely, and the paper is generally well organized. Bringing together ecological systems theory and network analysis is a meaningful step forward in the literature on child development and psychosocial adaptation.

The conceptualization of social adaptation is reasonable, though the manuscript would gain from a tighter alignment between how the construct is defined theoretically and how it is actually operationalized through the five measured dimensions. The case for using network analysis is made, but the criticism of traditional latent variable approaches feels somewhat rushed and would be more convincing with a deeper theoretical discussion.

Overall, the manuscript rests on solid conceptual foundations, though some sections need sharper and more precise argumentation.

The writing is mostly clear, but the Discussion section suffers from repetition, with similar points coming up more than once. The transitions between sections could also be smoother to improve the overall flow of the argument.

As for the statistical choices, the methods are appropriate and technically sound, but a few issues deserve attention: the use of mean imputation for missing data risks introducing bias and needs a stronger justification; and the description of the network comparison tests is too brief to allow replication. More transparency in the analytical workflow is needed before the study can be considered fully reproducible.

Research objectives are stated clearly, but the manuscript falls short when it comes to formulating explicit, testable hypotheses. Spelling out specific predictions that map onto the analytical strategy would considerably strengthen the paper.

The results are presented in a coherent and organized way. Network visualizations and centrality metrics are used appropriately, but some findings are labeled as "strong" or "important" without enough supporting context. Given the correlational nature of the data, interpretations should be worded more cautiously.

The figures are well chosen and serve the study's purposes. Network graphs and heat maps are fitting for this type of analysis, but the captions need to be expanded to help readers follow the results more independently.

The reference list is broad and largely up to date, with many recent and pertinent citations. However, this section also contains several formatting inconsistencies and typographical errors that need to be addressed before resubmission.

The conclusions follow reasonably from the findings, and the manuscript rightly draws attention to the role of interpersonal and learning adaptation. That said, some claims push slightly beyond what a cross-sectional design can actually support. The limitations are mentioned, but they would carry more weight if they were woven more directly into the interpretation of the results.

Finally, while the general analytical approach is described, the level of detail provided is not sufficient for full reproducibility. The authors should include more specific information on data preprocessing steps, parameter settings, and the software tools used, to meet current standards of scientific transparency.

Overall Recommendation: Reconsider after Major Revision

The manuscript addresses an important research question and employs an innovative analytical approach. However, revisions are required to strengthen theoretical justification, clarify methodological procedures, improve transparency, and refine the interpretation of results. With these improvements, the study could be a valuable contribution to the topic.

Author Response

Comments 1: The conceptualization of social adaptation is reasonable, though the manuscript would gain from a tighter alignment between how the construct is defined theoretically and how it is actually operationalized through the five measured dimensions. The case for using network analysis is made, but the criticism of traditional latent variable approaches feels somewhat rushed and would be more convincing with a deeper theoretical discussion.

Response 1: We sincerely thank the reviewer for this insightful and constructive feedback. In response to the suggestion, We have strengthened the theoretical description for determining the five measurement dimensions of social adaptation and refined the introduction of the shortcomings of traditional latent variable methods. Additionally, we have reorganized the relevant paragraphs to improve logical coherence and narrative flow (see Section 1, Page 2 paragraph 2 and 4, Lines69–76, ). We have updated the relevant text to read as follows:

Social adaptation refers to the ongoing adjustment of an individual's physical and mental state within dynamic social contexts, with the aim of achieving alignment with the current environment (Peng & Dai, 2023). According to the model of social competence, individuals acquire age-appropriate behaviors while living through continuously changing fields(Greenspan & Granfield, 1992); for school-age children, this adaptation should encompass several dimensions, including academic adaptation, life adaptation, emotional adaptation, interpersonal adaptation, and cognitive adaptation.

However, the former orientation focuses only on a single dimension of adaptation and fails to adequately account for the interactions among different dimensions. The latter relies on traditional latent variable models, which assume that the latent variable serves as the common cause of observed variables within the same dimension, and that all correlations among observed variables are fully explained by the latent variable. That is, it assumes that all dimensions are equally influenced by the same latent variable, thereby overlooking the possibility that different dimensions may vary in importance. Consequently, neither research orientation can effectively identify the key dimensions of social adaptation in the developmental process of left-behind children.

Comments 2: The writing is mostly clear, but the Discussion section suffers from repetition, with similar points coming up more than once. The transitions between sections could also be smoother to improve the overall flow of the argument.

Response 2: We sincerely thank the reviewer for this insightful comment. We agree that the Discussion section contained some repetitive statements and that transitions between sections could be improved. To address these issues, we have carefully revised the Discussion by removing redundant content, consolidating similar points, and reorganizing the paragraph structure (Section4.2, Page13, Line 424-439).

Furthermore, this study found that the connection between interpersonal adaptation(IA) and adaptation to life(AL) showed the strongest correlation in all three SA networks, followed by the connection between interpersonal adaptation(IA) and positive emotional adaptation(PEA). These results indicate a significant bidirectional relationship between interpersonal adaptation, learning adaptation, and positive emotional adaptation, which supports the idea of predictive relationships in the developmental cascade theory (Masten et al., 2005). Consistent results have also been found in college students by Liu and Li (2024).

We also examined the network variations in the SA network between rural left‑behind children and non‑rural left‑behind children. The findings revealed no significant differences in network structure invariance and global strength invariance between the two groups, but a notable distinction in edge invariance. Specifically, the connections between nodes in the SA network of rural left‑behind children were weak, indicating a lack of robust interactions across various dimensions of social adaptation within this group. In line with the developmental cascade theory of social adaptation, positive development in one dimension should positively influence other dimensions of adaptation (Masten & Cicchetti, 2010). Based on the findings of this study, this highlights that the limited interconnection among different social adaptation dimensions in left‑behind children impedes the establishment of a positive development cycle.

Comments 3: As for the statistical choices, the methods are appropriate and technically sound, but a few issues deserve attention: the use of mean imputation for missing data risks introducing bias and needs a stronger justification; and the description of the network comparison tests is too brief to allow replication. More transparency in the analytical workflow is needed before the study can be considered fully reproducible.

Response 3: We sincerely thank the reviewer for this insightful and constructive feedback.The data cleaning process involved excluding participants with >10% missing data, followed by verification that all individual items had <0.5% missing values. Given the low missing rate and sample homogeneity, mean imputation was employed based on established methodological recommendations. We have now explicitly detailed this procedure in the Statistical Analyses section. Accordingly, we have added a dedicated paragraph in the Discussion to highlight the practical implications of our study (see Section 2.4, Page 4, paragraph 5, Lines 182–185 ). We have updated the relevant text to read as follows:

In this on-site questionnaire study, participants with over 10% missing item-level data were excluded; all remaining items had missing rates below 0.5%, and given the homogeneous sample and low missing proportion, mean imputation was considered appropriate (Shan et al., 2023). In this study, SPSS version 25.0 was used for descriptive statistical analysis, internal consistency reliability test and correlation analysis, and R 4.3.1 in RStudio 1.2.5033 was was used used to estimate PS-SA network structure and network comparison. The network analysis approach follows the standard guidelines published by Epskamp (Epskamp et al., 2018).

Comments 4: Research objectives are stated clearly, but the manuscript falls short when it comes to formulating explicit, testable hypotheses. Spelling out specific predictions that map onto the analytical strategy would considerably strengthen the paper.

Response 4: Thank you very much for your thoughtful and constructive comments on our manuscript. We greatly appreciate your recognition that the research objectives are clearly stated. Regarding your concern about the lack of explicit, testable hypotheses, we have carefully revised the manuscript to address this issue. (Section 1, Page 3, Paragraph 1 and 4, Lines 93–95 and 126-129) . We have updated the relevant text to read as follows:

Specifically, this approach conceptualizes various dimensions of social adaptation as network nodes, with the lines between nodes representing the relationships among these dimensions. The thickness of each line indicates the strength of the correlation between nodes, while node centrality metrics reflect the importance of each dimension within the network (Opsahl et al., 2010). Network analysis emphasizes the interconnections among the different dimensions of social adaptation, using statistical modeling to highlight the most significant ones. This method effectively compensates for the limitations of traditional latent variable models, which often struggle to capture the relational patterns among social adaptation dimensions (CAI et al., 2010). Consequently, it provides a novel perspective for identifying key adaptive dimensions in children's development (Borsboom, 2017). Therefore, the first objective of this study is to construct a network structure of social adaptation and to compare the differences in this network structure between left-behind children and non-left-behind children.

Secondly, while prior studies have predominantly relied on regression analysis to examine the influence of parenting styles on social adaptation, they often overlook the specific mechanisms through which parenting practices directly affect adaptive outcomes (López-Soler et al., 2009; Schoeps et al., 2020). The use of network analysis can provide deeper insight into how parenting styles shape social adaptation and aid in developing targeted intervention strategies (Borsboom, 2017). Therefore, the second objective of this study is to construct a network structure of parenting styles and social adaptation, and to compare the differences in this parenting style–social adaptation network between left‑behind and non‑left‑behind children.

Comments 5: The results are presented in a coherent and organized way. Network visualizations and centrality metrics are used appropriately, but some findings are labeled as "strong" or "important" without enough supporting context. Given the correlational nature of the data, interpretations should be worded more cautiously.

Response 5: We appreciate your attention to detail. we have carefully revised the manuscript to address this issue. (Section3.3.1, Page 9, Paragraph 2 and 4, Lines321–23；) . We have updated the relevant text to read as follows:

Given the marked differences in parenting styles and social adaptation between left-behind and non-left-behind children, this study further examined variations in their network structures. We constructed three regularized partial correlation networks—for left-behind children (Fig. 4a), non-left-behind children (Fig. 4b), and the overall sample (Fig. 4c)—incorporating all 11 dimensions, resulting in 55 possible edges. The overall and non-left-behind networks each contained 29 non-zero edges with a mean weight of 0.05, while the left-behind children’s network had 32 non-zero edges and a higher mean weight of 0.06. Notably, the strongest undirected edge connecting the parenting styles and social adaptation sub-network in the left-behind children network was between IA and MEW (edge weight = 0.12). In contrast, the non-left-behind children network exhibited the strongest edge between LA and MEW (edge weight = 0.13), followed by the edge between PEA and MEW (edge weight = 0.11).

Comments 6: The figures are well chosen and serve the study's purposes. Network graphs and heat maps are fitting for this type of analysis, but the captions need to be expanded to help readers follow the results more independently.

Response6: We appreciate your attention to detail. We appreciate your attention to detail. we have carefully revised the manuscript to address this issue. (Section3, Page 7/8/10/11, Lines 260-261, 276, 286-288, 324–325) .

Figure 1. Heat map correlations. (1)The lower triangular matrix displays the correlation coefficients, with each value indicating the Spearman's c correlation coefficient. Only those coefficients that are statistically significant (p < 0.01) are included in the lower triangular matrix. (2)The upper triangular matrix illustrates the significance levels, *p<0.05,**p<0.01,***p<0.001, while the empty boxes indicate coefficients do not survive this correction.

Figure 2. Network structure of social adaptation. Notes: (1) A, B, and C respectively represent the social adaptation network structure of Overall sample of children, left-behind children, and non-left-behind children. (2)Each node represents a dimension of social adaptation.(3) Green lines mean positive connections, and the edge thickness indicates correlation strength. (4) To facilitate visual comparisons of the network structures across different groups, this study employs the average layout function, ensuring that the positions of identical nodes remain consistent.

Figure 3. Centrality plots for three SA networks. The X-axis represents the centrality indices as standardized z-scores, indicating that a higher estimate corresponds to greater centrality of the item, while the Y-axis displays the 11 variables.

Figure 4. Network structure of parenting styles-social adaptation. Notes: (1)a, b, and c respectively represent the social adaptation network structure of children as a whole, left-behind children, and non-left-behind children. (2)Each node represents a dimension of parenting styles and social adaptation. (3)Green lines mean positive connections, red lines mean negative connections, and the edge thickness indicates correlation strength. (4)To facilitate visual comparisons of the network structures across different groups, this study employs the average layout function, ensuring that the positions of identical nodes remain consistent.

Comments 7: The reference list is broad and largely up to date, with many recent and pertinent citations. However, this section also contains several formatting inconsistencies and typographical errors that need to be addressed before resubmission.

Response7: Thank you for pointing this out. We have thoroughly reviewed the citation style throughout the manuscript and have revised both in-text references and the reference list to align with the formatting requirements of Behavioral Sciences. We appreciate your attention to detail.

Comments 8: Finally, while the general analytical approach is described, the level of detail provided is not sufficient for full reproducibility. The authors should include more specific information on data preprocessing steps, parameter settings, and the software tools used, to meet current standards of scientific transparency.

Response8: Thank you for this constructive suggestion. We have revised the manuscript to include more detailed descriptions of data preprocessing steps, parameter settings, and software tools(Section2.4, Page5, Lines 194-196, 204-207, 214-218,225-226 to ensure full reproducibility. We have updated the relevant text to read as follows:

To assess node importance in the network, we computed Expected Influence (EI) and Bridge Expected Influence (BEI) using the R packages “qgraph” and “networktools”, identifying central and bridge nodes (Epskamp et al., 2018; Jones & McNally, 2021). EI was chosen for its better performance in networks with both positive and negative edges compared to other centrality indices (McNally, 2021). High EI reflects stronger positive associations with other nodes and greater overall network influence (Jones & McNally, 2021; Robinaugh et al., 2016). High-BEI nodes serve as crucial connectors between symptom clusters and may facilitate the development or persistence of other symptoms or disorders (Jones & McNally, 2021).

The evaluation of network stability and accuracy was conducted utilizing the “bootnet” package (Epskamp et al., 2018). We executed 1000 bootstrap samples to estimate the confidence intervals (CIs) at 95% for edge weights and examined the differences among edges and nodes. A narrower CI at 95% reflects greater precision in the edge weights within the network. The stability of the node centrality indices was evaluated through the case-dropping bootstrap method, which indicates the highest permissible percentage of data that may be omitted without jeopardizing stability. The Correlation Stability Coefficient (CS) was computed, where a CS score of ≥0.25 is considered satisfactory, while values exceeding 0.5 are regarded as preferable (Epskamp et al., 2018).

The “NetworkComparisonTest” was used to compare the networks of rural left-behind children and non-rural left-behind children. Three tests were conducted to evaluate these differences: a test for network structure invariance, a test for global strength invariance, and a test for edge strength invariance (Van et al., 2023). The network structure invariance test assessed variations in the strength of the maximum edge within the network; the global strength invariance test assessed differences in the total edge strength; and the edge strength invariance test examined variances in specific edges within the network.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have addressed a very interesting topic. Their choice of methodology is also commendable.

Nevertheless, the literature review could benefit from more recent research findings and a broader international scope, with greater emphasis on cultural influences, differences, opportunities, and obstacles. The description of the sample and methods is accurate, but it would be helpful to provide examples for the factors (if the questionnaire authors permit this).

Figures 1 and 3 are very difficult to read; please enlarge the data.

The data analysis is accurate. After expanding and clarifying the theory, you may also refine the interpretation. It would be worthwhile to formulate conclusions from a practical perspective.

Congratulations on the work.

Author Response

Comments 1: Nevertheless, the literature review could benefit from more recent research findings and a broader international scope, with greater emphasis on cultural influences, differences, opportunities, and obstacles. The description of the sample and methods is accurate, but it would be helpful to provide examples for the factors (if the questionnaire authors permit this).

Response 1: Thank you for the helpful comments. We have (1) updated the literature review with recent international studies and a discussion of cultural factors(see Section 1, Page 2 , Lines 64–67), and (2) added factor examples in the Methods section (with permission from the questionnaire authors). Revisions are highlighted in the manuscript.(see Section 2.3, Page 4, Lines 157-161 and 171–174). We have updated the relevant text to read as follows:

Presently, research on the social adaptation of left-behind children primarily exists in two orientations. The first orientation categorizes left-behind children into heterogeneous groups based on their response patterns in specific dimensions of social adaptation, further exploring the differential patterns among these categories (Zhao et al., 2019; Han et al., 2025). For instance, Li et al. employed latent class analysis to classify left-behind children into three types: well-adapted, those facing adaptation difficulties, and those exhibiting behavioral impulsivity, based on emotional and behavioral issues(Zhi-Hua et al., 2014). The study identified significant differences among the categories concerning gender and grade. The second research orientation evaluates the adaptation levels of left-behind children through a comprehensive score of social adaptation, enabling a comparison of adaptation differences among various groups (Fan et al., 2009; Hou et al., 2014). For instance, Fellmeth et al.(2018) conducted a meta-analysis comparing the overall social adaptation levels of left-behind and non-left-behind children(NLBC) across 111 studies, revealing that left-behind children generally exhibit more anxiety, depression, and other maladjustment problems.

The social adaptation of rural children was assessed with the Rural Children Social Adaptation Questionnaire (RCSAQ; Dai, 2019).The scale includes five dimensions: interpersonal adaptation (e.g.”Many classmates in my class are very kind to me”), learning adaptation (e.g.”I sincerely love learning”), positive emotional adaptation (e.g.”I can manage my negative emotions well”), cognitive adaptation (e.g.”I think it is normal for classmates to often argue about certain issues”) and life adaptation (e.g.“At home, I regularly do chores”), comprising a total of 26 items on a 5-point Likert scale ranging from 1 (not like me at all) to 5 (completely like me). Each dimension represents a positive aspect, and higher scores indicate a higher level of social adaptation for the rural children. This scale has demonstrated good reliability and validity in previous studies with Chinese children (Peng & Dai, 2023). The Cronbach's α coefficient for the scale was 0.89.

Parenting styles was assessed with Chinese version of the short-form Egna Minnen av Barndoms Uppfostran (s-EMBU; Jiang et al., 2010). The s-EMBU consisted of two subscales: one for fathers and one for mothers. Each subscale is further divided into three dimensions: emotional warmth (e.g.”My parents always try to encourage me to be the best”), rejection(e.g.”My parents often treat me in an embarrassing way”), and over-protection(e.g.”My parents forbid me from doing things that other kids can do because they're worried I might get hurt”). Emotional warmth is considered a positive dimension, while rejection and over-protection are viewed as negative dimensions. In total, the overall scale comprises a total of 42 items, with each subscale contains 21 items that are identical in content. Each item was rated on a 4-point Likert scale, ranging from 1 = strongly disagree to 4 = strongly agree. A higher score indicates a higher tendency for the child's father or mother to exhibit this particular parenting style. This scale has demonstrated good reliability and validity in previous studies with Chinese children (Song et al., 2024). The Cronbach's α coefficient for the scale was 0.91.

Comments 2: Figures 1 and 3 are very difficult to read; please enlarge the data.

Response 2: Thank you for pointing this out. We have enlarged the data in Figures 1 and 3 to improve readability. The updated figures are now included in the revised manuscript.

Comments 3: The data analysis is accurate. After expanding and clarifying the theory, you may also refine the interpretation. It would be worthwhile to formulate conclusions from a practical perspective.

Response 3: Thank you for the helpful comments. We have expanded and clarified the theoretical basis, refined the interpretation accordingly, and added practical conclusions. Revisions are highlighted in the manuscript(see Section 5, Page 12, paragraph 1, Lines 536–539). We have updated the relevant text to read as follows:

To the best of our knowledge, this is the first study to utilize network analysis technology in comparing the social adaptation network structural characteristics of left-behind children and non-left-behind children, as well as to examine the influence of parental parenting styles on social adaptation. This study enhances the understanding of the interrelationship between parental rearing styles and the social adjustment . First, within the SA network, we found that the core dimensions of left-behind children and their non-left-behind counterparts are identical, with both groups exhibiting IA and LA. This suggests that these two dimensions serve as the most effective indicators of children's social adaptation. A comparison of the networks of both groups revealed similar network structures and global strengths, yet notable differences in specific dimension associations. Second, within the PS-SA network, we found that the core dimensions of left-behind children differ from those of non-left-behind children. Specifically, interpersonal adaptation(IA), learning adaptation(LA), and paternal over protection(PO) comprise the core dimensions of the left-behind children network, whereas, interpersonal adaptation(IA), learning adaptation(LA), and positive emotional adaptation(PEA) constitute the core dimensions of the non-left-behind children network. Although the networks of both groups displayed similar structures, they varied in global strength and specific dimension associations. The research findings tell us that social adaptation interventions for left-behind children should prioritize improving their interpersonal and learning adaptation, along with reducing paternal overprotection, to foster a more positive developmental trajectory.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Review: behavsci-4312525

Thank you for the opportunity to revise the manuscript „ The relationship between social adaptation and parenting styles in left-behind and non-left-behind children: a network analysis”. The study addresses an important and timely topic in child development and family psychology.

Abstract: In the conclusion, the statement that the findings provide “precise targets for adaptation education” for parents and schools should be moderated, as it overstates the level of practical and intervention-specific precision that can be inferred from the study.

Introduction: The introduction is generally well structured and provides a clear overview of the issue of social adaptation in children separated from their parents. The authors successfully present the importance of the topic, relevant empirical findings, as well as the limitations of previous research approaches, which leads to a clear justification for the use of network analysis.

However, some parts of the introduction use language that may imply causal mechanisms (e.g., “shape”, “directly affect”), which may be overstated given the likely cross-sectional design and the correlational nature of network analysis. A more cautious wording emphasizing associations and network connections would improve methodological precision.

Methodology: In the network interpretation section, psychopathology-related terminology (e.g., “symptoms,” “symptom clusters,” “disorders”) is used, although the nodes in this study represent dimensions of social adaptation and parenting styles. It is recommended to consistently use terms such as “dimensions” or “variables” to avoid an unintended psychopathological framing.

Discussion: The interpretation of the results occasionally exceeds the methodological scope of network analysis, as associative relationships are framed in causal and developmental terms (e.g., “core dimensions played a critical role,” “these results indicate a significant bidirectional relationship,” and “this highlights that the limited interconnection… impedes the establishment of a positive development cycle”), and should be more cautiously restricted to describing associations between dimensions.

Author Response

Comments 1: In the conclusion, the statement that the findings provide “precise targets for adaptation education” for parents and schools should be moderated, as it overstates the level of practical and intervention-specific precision that can be inferred from the study.

Response 1: Thank you for this careful comment. We agree that the term “precise targets” overstates the practical implications. In the revised manuscript, we have replaced it with “potential targets” to better reflect the exploratory nature of our findings. (see Section Abstract, Page1, Lines 77–81 and 27–30). We have updated the relevant text to read as follows:

Conclusions: these findings visualise and specify how distinct parenting dimensions relate to different facets of social adaptation, offering parents and schools potential targets for adaptation education tailored to left‑behind children.

Comments 2: some parts of the introduction use language that may imply causal mechanisms (e.g., “shape”, “directly affect”), which may be overstated given the likely cross-sectional design and the correlational nature of network analysis. A more cautious wording emphasizing associations and network connections would improve methodological precision.

Response 2: Thank you for this important methodological comment. We agree that the wording in the introduction may have inadvertently implied causal relationships. We have modified the relevant statements in the manuscript as follows(Section 1, Page 3, Line 105-108 and 121-126).

Moreover, according to ecological systems theory, the family constitutes the most critical component within the micro-environmental system (Bronfenbrenner, 1986). Within the family context, parenting style serves as a central factor significantly influencing the social adaptation of left-behind children (Fellmeth et al., 2018). Research indicates that positive parenting practices—such as emotional warmth—are associated with reduced anxiety, depression, loneliness, and behavioral problems, while also promoting greater happiness among these children (Carlo et al., 2011; Khaleque, 2013; Butterfield et al., 2021; Zhang, 2025 ) . In contrast, negative parenting styles marked by rejection or over protection have been correlated with poorer social adaptation outcomes (Cong et al., 2020; Schiff et al., 2021; Wu & Su, 2024). Furthermore, studies suggest that parenting styles not only directly affect the social adaptation of left‑behind children but also indirectly predict it through mediating variables such as self‑control and parent‑child conflict (Pan et al., 2021; Zhang et al., 2023).

Comments 3: In the network interpretation section, psychopathology-related terminology (e.g., “symptoms,” “symptom clusters,” “disorders”) is used, although the nodes in this study represent dimensions of social adaptation and parenting styles. It is recommended to consistently use terms such as “dimensions” or “variables” to avoid an unintended psychopathological framing.

Response 3: Thank you for this careful observation. We agree that the use of psychopathology-related terminology (e.g., “symptoms,” “symptom clusters,” “disorders”) was inappropriate for our nodes, which represent dimensions of social adaptation and parenting styles. Accordingly, we have revised the manuscript by replacing these terms with more neutral ones such as “dimensions” or “variables.” A total of five such replacements have been made in the network interpretation section. We believe these changes eliminate the unintended psychopathological framing and improve the conceptual clarity of the paper.

Comments 4: The interpretation of the results occasionally exceeds the methodological scope of network analysis, as associative relationships are framed in causal and developmental terms (e.g., “core dimensions played a critical role,” “these results indicate a significant bidirectional relationship,” and “this highlights that the limited interconnection… impedes the establishment of a positive development cycle”), and should be more cautiously restricted to describing associations between dimensions.

Response 4: We sincerely thank the reviewer for this insightful and constructive feedback. We have revised the interpretations to avoid causal or developmental language, limiting them to describing associations between dimensions. Changes are highlighted in the manuscript(Section 4, Page1 3, Line 412-416 and 424-439).

Network analysis results reveal that interpersonal adaptation(IA) and learning adaptation(LA) were the core dimensions in three SA network structures, highlighting the importance of interpersonal adaptation and learning adaptation for both groups of children. These dimensions may play an important role in social adaptation and can have implications for various other areas. Previous studies have found that children who struggle with interpersonal or academic adjustment may experience challenges in emotional well-being, life adjustments, and may exhibit maladaptive behaviors (Moilanen et al., 2010). Specifically, those with poor interpersonal adjustment may exhibit feelings of loneliness, social withdrawal, and aggression (Jin et al., 2023; Chen et al., 2015), while those with academic struggles may face internalizing issues like self-doubt, anxiety, and depression (Deighton et al., 2018). These difficulties can hinder their ability to effectively adapt to their current social environment.

These results indicate a significant bidirectional relationship between interpersonal adaptation, learning adaptation, and positive emotional adaptation, which supports the idea of predictive relationships in the developmental cascade theory (Masten et al., 2005). Consistent results have also been found in college students by Liu and Li (2024).

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Authors have satisfactorily addressed the comments raised in the review and have revised the article to make it suitable for publication. In its current form, in my opinion, the article is ready for publication.