How to make models more useful

Barton, C. Michael; Lee, Allen; Janssen, Marco A.; Leeuw, Sander van der; Tucker, G. E.; Porter, Cheryl; Greenberg, Josh; Swantek, Laura; Frank, Karin; Chen, Min; Jagers, Bert

doi:10.1073/pnas.2202112119

Cited by 26 publications

(16 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition to enhancing model replication and documentation through the use of ODD or ODD+D protocols, researchers should also consider adopting “FAIR” principles (Findability, Accessibility, Interoperability, and Reusability) when organizing the code. For instance, researchers should ready their code by ensuring it is clean and thoroughly commented …”

Section: Discussionmentioning

confidence: 99%

“…For instance, researchers should ready their code by ensuring it is clean and thoroughly commented. 181 Studies have used the ODD protocol to improve the readability and reproductivity of the model. In the study of applying the microbial individuality model to simulate microbial reactions in the anaerobic digestion of sewage sludge, the authors generalized the microbial responses as the wolf-sheep predation model in ABM.…”

Section: Solutions To Replication and Documentation Issuementioning

confidence: 99%

See 1 more Smart Citation

Agent-Based Modeling in Water Science: From Macroscale to Microscale

Xu,

Hsu,

et al. 2024

ACS EST Water

View full text Add to dashboard Cite

Agent-based modeling (ABM) has been widely applied in water science, covering a wide range of research topics, from evaluating the efficiency of water management strategies at the macroscale to simulating microbial and algal colony reactions at the microscale. The capability of ABM for simulating how system performance is affected by individuals' behaviors and reactions makes it an ideal tool for representing complex adaptive systems in the real world. This paper provides a comprehensive review of the application of ABM in water science and identifies the current common problems. Several potential solutions to the problems were discussed, including (1) employing standard methods for validation, (2) adopting the ODD (Overview, Design, and Details) protocol and ODD+D (ODD + Decision) protocol to enhance readability and reproducibility of the model, and (3) applying a scaling-down approach and superindividual approach to reduce model complexity and facilitate computational efficiency.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Solutions To Replication and Documentation Issuementioning

confidence: 99%

Agent-Based Modeling in Water Science: From Macroscale to Microscale

Xu,

Hsu,

et al. 2024

ACS EST Water

View full text Add to dashboard Cite

show abstract

“…Such models would make theories more rigorous and complete, and could highlight questions for future research. Following the recommendations of Barton et al (2022), these models should be easily reproducible, with publicly available code that is accessible to those with minimal modeling experience, allowing them to be adapted by other researchers who wish to develop their own hypotheses. As noted, it is also important that future models can simulate performance on common creative tasks, to allow models to be readily compared to both human data and the performance of other models.…”

Section: Toward Greater Integration Between Ncr and Computational Mod...mentioning

confidence: 99%

“…To further demonstrate how a creative task can be modeled computationally, we have included an implementation of this simple model in MATLAB code in the online supplemental materials, together with a detailed overview. We have made the code accessible to researchers with minimal modeling experience, implementing a major recommendation made by Barton et al (2022) to enhance the usability of computational models. Of note, this toy model is by no means intended as a definitive model of creativity, but simply as an example of how verbal theories of creative task performance can be translated into formal models for researchers who may have limited or no prior experience in computational modeling.…”

Section: The Benefits Of Modelingmentioning

confidence: 99%

Toward greater computational modeling in neurocognitive creativity research.

Lloyd-Cox,

Pickering,

Beaty

et al. 2023

Psychology of Aesthetics, Creativity, and the Arts

View full text Add to dashboard Cite

Creative cognition is the driving force behind all cultural and scientific progress. In recent years, the field of neurocognitive creativity research (NCR) has made considerable progress in revealing the neural and psychological correlates of creative cognition. However, a detailed understanding of how cognitive processes produce creative ideas, and how these processes interact differently across tasks and individuals, remains elusive. In this article, we argue that the increased adoption of computational modeling can help greatly in achieving this goal. While the verbal theories guiding NCR have evolved from broader accounts into more specific descriptions of neurocognitive processes, they remain more open to interpretation and harder to falsify than formal models. Translating theories into computational models can make them more concrete, accessible, and easier to compare, and helps researchers to develop causal hypotheses for how variation in cognitive factors leads to variation in creative outcomes. Currently, however, computational modeling of creativity is conducted almost entirely separately from NCR, and few attempts have been made to embody the cognitive theories of NCR in models that can simulate performance on common lab-based tasks. In this article, we discuss theories of creative cognition and how they might benefit from the wider adoption of formal modeling. We also examine recent computational models of creativity and how these might be improved and better integrated with NCR. Finally, we describe a pathway toward a mechanistic understanding of creative cognition through the integration of computational modeling, psychological theory, and empirical research, outlining an example model based on dual-process accounts. Public Significance StatementThis review argues that creativity research would benefit greatly from the wider adoption of computational modeling. We discuss how translating verbal theories of creative cognition into formal computational models can make them more rigorous, accessible, and communicable, and can highlight questions for future research. We examine previous models of creativity and explain how these can be improved to benefit our understanding of human creative cognition and the development of artificial creative systems.

show abstract

“…13,14 Even widely used scales like the Center for Epidemiologic Studies Depression Scale (CESD-10) are interpreted as dichotomous variables 15 or continuous variables, 16 leading to the utilization of different statistical models. 17,18 Similarly, there are different approaches to operationalizing grandparenting. After treating grandparenting as a binary variable based on whether grandparents provide care to grandchildren, 19,20 some scholars have categorized it based on whether the caregiving duration is at least 15 h per week.…”

Section: Introductionmentioning

confidence: 99%

Revisiting the association between grandparenting and mental health in China: New evidence from the harmonized CHARLS

Wang,

Xiao,

et al. 2024

Int J Geriat Psychiatry

View full text Add to dashboard Cite

BackgroundIn the context of the global aging challenge, an increasing number of middle‐aged and older adults (MAOAs) are engaging in grandparenting. However, the effect of grandparenting on the mental health of caregivers has shown inconsistent findings. To effectively promote healthy aging, it is imperative to adopt a comprehensive perspective and employ a rigorous approach to further investigate the relationship between these two social phenomena.MethodsThe data from the Harmonized China Health and Retirement Longitudinal Study were analyzed, focusing on MAOAs with at least one grandchild. Mental health assessments used the center for epidemiologic studies depression scale scale. The study employed a series of difference‐in‐differences (DID) models, especially complemented by propensity score matching, to evaluate the average treatment effect for the treated (ATT) on mental health of caregivers, considering covariates like personal and family characteristics. The intervention perspective includes both the provision and cessation of grandparenting.ResultsThe study found that providing grandchildren care does not have a significant effect on the mental health of grandparents, in comparison to those who have never engaged in such care (ATT = −0.172, T = 0.65, p = 0.517 in the PSM‐DID model). Furthermore, ceasing this care also appears to have no substantial effect on the mental health of the caregivers, relative to individuals who have consistently offered grandchildren care (ATT = 0.060, T = 0.26, p = 0.795 in the PSM‐DID model). Furthermore, subsequent robustness analyses consistently supported these findings, even when considering data from different survey waves.ConclusionsIn contrast to many prior studies that have reported either positive or negative effects, our research reveals that grandparenting exerts no significant effect on the mental health of MAOAs. Consequently, health practitioners and policymakers should carefully consider the diverse cultural context when tailoring interventions and support strategies.

show abstract

How to make models more useful

Cited by 26 publications

References 18 publications

Agent-Based Modeling in Water Science: From Macroscale to Microscale

Agent-Based Modeling in Water Science: From Macroscale to Microscale

Toward greater computational modeling in neurocognitive creativity research.

Revisiting the association between grandparenting and mental health in China: New evidence from the harmonized CHARLS

Contact Info

Product

Resources

About