Relational Reasoning Across Cultures: Consistency and Variation in South Korea, the United States, and China

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is an interesting manuscript that addresses an important question in cross-cultural cognitive development using a rich and carefully designed multi-experiment approach. The work is clearly motivated and the empirical contribution is valuable. However, several aspects of the manuscript would benefit from substantial revision to improve clarity, strengthen the theoretical framing, and align the interpretation more closely with the strength of the evidence.  

• Across the Discussion, there are several instances where the results are interpreted in strongly causal terms (e.g., environmental input “shaping priors” or cognitive biases “playing a causal role” in developmental differences). Given that the study is observational and cross-sectional across cultures, these claims go beyond what the design can support. I would encourage the authors to consistently soften causal language and instead frame conclusions in terms of associations or compatibility with theoretical predictions. This will preserve the contribution without overextending the inferential scope.

• The interpretation of Experiment 2 as evidence for a “neutral bias” or “absence of bias” in Korean 3 year old seems too strong given the forced-choice design and the Bayesian evidence reported (BFs close to 1). The data are more appropriately described as inconclusive with respect to systematic bias. It would also be helpful to acknowledge more explicitly that the task may have limited sensitivity to detect weak or graded preferences at this developmental stage. 

• In Experiment 3, a Bayesian logistic regression is initially reported but later replaced by a frequentist model due to convergence issues. This is not fully explained, and it raises questions about model robustness. I would recommend: clarifying why convergence failed, and reporting any diagnostics attempted.

• The interpretation of cross-cultural differences sometimes moves quickly from descriptive differences to broad explanatory claims about cultural learning environments. While this is theoretically interesting, the current design does not allow strong inference about specific cultural mechanisms. It would strengthen the manuscript to more explicitly acknowledge alternative explanations (e.g., sampling differences, task familiarity, or educational contexts) and to avoid broad generalizations beyond the sampled populations.

•  

  If helpful, the manuscript would benefit substantially from integrating more recent empirical literature to directly support each of the main claims. Updating this aspect would significantly strengthen the theoretical robustness of the paper and reduce vulnerability to reviewer concerns regarding temporal validity.

With revisions addressing these points, the paper has the potential to make a meaningful contribution to the literature on cross-cultural cognitive development.

 

I recommend major revision due to concerns in three main areas:

The manuscript frequently draws causal-level interpretations from cross-sectional and observational cross-cultural data, particularly regarding environmental input shaping cognitive biases. These claims exceed what the design can support and require substantial moderation.

The transition from Bayesian to frequentist modeling due to convergence issues in Experiment 3 is insufficiently explained. Model robustness and justification for analytic choices need to be clarified.

Experiment 2 is interpreted as evidence for a “neutral bias” in Korean children despite Bayes factors indicating inconclusive evidence. This interpretation is currently too strong and should be reframed more cautiously.

The manuscript relies heavily on older foundational literature (e.g., Tardif et al., 1997; Zhang et al., 2005; Shin & Choi, 2008) to support claims about contemporary cultural and linguistic environments. Given the paper’s strong claims about sociocultural variability and change, this literature base should be substantially updated to include recent empirical work (2018–2025).

While the empirical contribution is strong and the cross-cultural dataset is valuable, the manuscript currently requires revision to align interpretation with evidential strength and to modernize its theoretical grounding. I believe these issues are fully addressable and would significantly improve the paper’s impact and rigor.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript presents a well-designed cross-cultural investigation of the development of relational reasoning. Drawing previous work comparing children from the United States and China, the study introduces a third cultural context, South Korea, to evaluate competing predictions regarding the effects of linguistic and cultural experience on the development of relational reasoning. The experiments are methodologically rigorous, the analyses are appropriate, and the findings are novel and theoretically interesting.

The principal contribution of the paper lies in demonstrating that developmental trajectories of relational reasoning are neither universal nor simply dichotomous. Instead, the Korean data reveal an intermediate developmental pattern that challenges simple East-West distinctions and supports the broader claim that environmental experience shapes the deployment of relational reasoning.

Overall, the manuscript deserves publication. However, its theoretical contribution could become stronger if the findings were interpreted within a broader developmental architecture of cognition. Below I discuss the strengths and weaknesses of the paper.

Strengths

The paper makes a significant empirical contribution by introducing South Korea as a theoretically informative comparison between the United States and China. Rather than merely replicating previous findings, the authors identify a distinct developmental trajectory deviating from the American and the Chinese pattern. This enriches the literature on cross-cultural cognitive development.

The three experiments are well connected, addressing early emergence, developmental bias, and developmental change. The Bayesian analyses are appropriate, the methodological procedures replicate previous work, and the authors discuss the limitations of their interpretations.

One of the manuscript's greatest strengths is its developmental orientation. Rather than treating relational reasoning as a fixed cognitive capacity, the authors examine how its expression changes over development. Particularly important is the evidence suggesting that young children possess relational competence before the emergence of later biases, followed by temporary decline and subsequent recovery. This developmental trajectory represents an important phenomenon that deserves further theoretical explanation.

Moving beyond simple cultural dichotomies. The paper successfully demonstrates that East Asian cultures cannot simply be grouped together under a single developmental profile. The Korean findings illustrate that relatively small differences in linguistic experience, parenting practices, educational systems, and attentional styles may produce measurable differences in developmental trajectories. This is an important contribution that moves the field beyond simplistic cultural comparisons.

Careful interpretation. The authors generally avoid over-interpreting the findings. They acknowledge that the present experiments cannot identify the causal mechanisms responsible for the observed cultural differences. This caution strengthens the credibility of the paper.

Major Theoretical Limitations

The theoretical interpretation of the study remains relatively narrow. The manuscript treats relational reasoning primarily as a behavioral outcome rather than as an emergent property of a developing cognitive architecture. Therefore, the findings raise important questions that remain unanswered.

Throughout the manuscript, relational reasoning is largely treated as a single cognitive capacity measured by performance on the causal relational match-to-sample task. However, successful performance almost certainly requires the coordinated operation of multiple cognitive processes, including search processes, abstraction of common structure, and representation of relational information, occurring in a context where attentional processes, working memory, and response monitoring and evaluation are important. These components are not distinguished. Therefore, it remains unclear which aspect of cognition changes in development or varies across cultures. The theoretical framework would therefore benefit from decomposing relational reasoning into its constituent cognitive processes.

The rational learner account explains developmental change primarily in terms of changing priors. This may be an elegant computational description, but it leaves an important developmental question unanswered: What changes inside the cognitive system that produces these changing priors? The observed developmental trajectory presumably reflects changes in representational organization, executive control, attentional allocation, abstraction, and learning mechanisms. Without specifying these processes, the explanation remains primarily descriptive rather than mechanistic.

Throughout the manuscript, culture, language, caregiver input, attentional style, and educational experience are discussed as possible explanatory factors. However, these constructs remain relatively broad and heterogeneous. The theoretical argument would become stronger if the manuscript specified the mechanisms through which environmental experience modifies cognitive development. For example, environmental influences might alter representational context, attentional priorities, executive control, abstraction processes, or learning strategies. Ignoring these mechanisms weakens the explanatory power of the framework.

The manuscript is basically disconnected from the cognitive development literature. It briefly discusses executive function but does not integrate relational reasoning into broader theories of cognitive development. Relational reasoning is unlikely to develop independently of executive control, working memory, abstraction, language, processing efficiency, or metacognitive regulation. Understanding how these systems interact would substantially deepen the developmental interpretation of the present findings.

In the fashion above, the manuscript is basically disconnected from the intelligence literature. It is a significant omission that the relationship between relational reasoning and the broader architecture of intelligence is ignored. Relational reasoning is one of the strongest cognitive predictors of psychometric intelligence. However, the present manuscript does not discuss whether the observed developmental trajectories may reflect broader developmental changes in general cognitive organization. Addressing this issue would substantially increase the theoretical importance of the findings beyond the relational reasoning literature.

Suggestions for Strengthening the Theoretical Contribution

Rather than viewing relational reasoning as an isolated competence, future work could treat it as the emergent outcome of multiple interacting cognitive systems, including representation, abstraction, executive control, attention, and working memory. Such an approach would allow researchers to explain why relational reasoning changes.

Future studies could simultaneously assess processes such as working memory, processing speed, attentional control, and abstraction, together with relational reasoning. This would permit direct evaluation of the mechanisms responsible for developmental and cultural variation.

Rather than treating culture as a direct cause of relational reasoning, future models could examine how cultural experience influences specific cognitive processes, which in turn shape relational reasoning. Such mediation models would provide considerably stronger theoretical explanations than direct culture-performance associations.

The observed U-shaped trajectory strongly suggests that developmental change is dynamic rather than monotonic. Future work would benefit from longitudinal or dynamic modeling approaches capable of capturing developmental reorganization rather than simple age differences.

Finally, future work could investigate how developmental changes in relational reasoning relate to the emergence of general cognitive ability. Doing so would situate the present findings within the broader literature on intelligence development and considerably increase their theoretical impact.

In short, this manuscript is a promising empirical contribution to the study of relational reasoning and cross-cultural cognitive development. The experiments are carefully executed, the findings are novel, and the conclusions are cautious. Its primary limitation is theoretical rather than empirical. The manuscript successfully documents developmental variation but provides only a partial explanation for why such variation emerges. The rational learner framework offers an elegant account in terms of changing priors, yet it does not fully specify the underlying developmental architecture that gives rise to those changing priors. Future work would benefit from embedding relational reasoning within a broader developmental framework in which abstraction, representation, executive control, attention, working memory, and language jointly contribute to the emergence of relational reasoning across development. Such an approach would not diminish the importance of the present findings; rather, it would provide a richer account of how environmental experience becomes translated into developmental change. Below I give several papers showing how relational reasoning is embedded into the general theory of intelligence and intellectual development that may be useful in putting this study into a broader developmental/psychometric perspective.

A technical note: The figures need a note that would enable the reader to grasp what is required by the children and what is the correct response.

Below I give some references that show how relational reasoning and its development is embedded in the theory of intelligence and intellectual development.

Demetriou, A., Makris, N., Spanoudis, G., Karousou, A., Kazi, S., Oikonomakou, D., Bikos, T. (2026). How profiles of general intelligence change with age: A comprehensive model.  Cognitive Development, 79 (2026) 101735, https://doi.org/10.1016/j.cogdev.2026.101735

Demetriou, A., Savva, A., & Spanoudis, G. (2025). SARA-C: A Core Mechanism Underlying g in Evolution and Development. Behavioral and Brain Science, https://doi.org/10.1017/S0140525X25100745.

Kazali, E., Spanoudis, G., & Demetriou, A. (2024). g: Formative, reflective, or both? Intelligence, 107, 101870.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Accept in present form

Reviewer 2 Report

Comments and Suggestions for Authors

The paper may be accepted in its present form.