Article
Public, Environmental & Occupational Health
Hiba N. Kouser, Ruby Barnard-Mayers, Eleanor Murray
Summary: Systems models are a natural choice for bridging the gap between social epidemiology and causal inference, and can improve our understanding of quantitative causal effects in social epidemiology, such as optimizing resource distribution during pandemics to reduce social inequalities.
JOURNAL OF EPIDEMIOLOGY AND COMMUNITY HEALTH
(2021)
Article
Psychology, Clinical
Tyler M. Moore, Benjamin B. Lahey
Summary: The study suggests that residualized lower order factors are correlated with bifactor-specific factors to the extent that factor loadings are proportional. When actual data violates proportionality, specific and residualized lower order factors are not always highly correlated and have different correlations with criterion variables, even if both models fit acceptably. This indicates that bifactor models should be preferred for hierarchical modeling due to the limitations of proportionality constraints in second-order models.
CLINICAL PSYCHOLOGICAL SCIENCE
(2022)
Article
Multidisciplinary Sciences
Jonathan Barrett, Robin Lorenz, Ognyan Oreshkov
Summary: Causal reasoning is crucial to science, but quantum theory poses challenges to it. While recent developments in intrinsically quantum causal models address some issues, extending these models to cyclic causal structures is necessary to fully tackle the challenges posed by quantum correlations and indefinite causal processes.
NATURE COMMUNICATIONS
(2021)
Review
Psychology, Clinical
Debbie Huang, Ezra Susser, Kara E. Rudolph, Katherine M. Keyes
Summary: The network paradigm considers psychiatric disorders as being caused by interconnected symptoms. However, studies on depression within this paradigm have not sufficiently addressed the necessary assumptions for causal inference. The lack of discussion and testing of causality assumptions undermines the credibility of depression network studies.
PSYCHOLOGICAL MEDICINE
(2023)
Article
Nutrition & Dietetics
Georgia D. Tomova, Mark S. Gilthorpe, Peter W. G. Tennant
Summary: This study explores and illustrates the theory and performance of the leave-one-out and energy partition models for estimating substitution effects in nutritional epidemiology. Through Monte Carlo data simulations, it was found that both models perform well when substituting a single exposure with a single component, but bias may occur when multiple components are involved in the substitution. Additionally, adjusting for total energy intake and considering foods measured in grams or portions may result in obscure estimands.
AMERICAN JOURNAL OF CLINICAL NUTRITION
(2022)
Article
Mathematical & Computational Biology
Rachel C. Nethery, Nina Katz-Christy, Marianthi-Anna Kioumourtzoglou, Robbie M. Parks, Andrea Schumacher, G. Brooke Anderson
Summary: Strategic preparedness can reduce the health impacts of tropical cyclones, but a more comprehensive understanding of cyclone epidemiology is needed. A machine learning approach can provide high-precision estimation of health impacts and identify the communities at highest risk.
Article
Psychology, Social
Sylia Wilson, Thomas M. Olino
Summary: A developmental perspective is essential for advancing the fields of personality science and psychopathology. Hierarchical models of individual differences can help delineate personality and psychopathology, while longitudinal studies are needed to understand their temporal ordering and dynamic associations over time.
JOURNAL OF PERSONALITY
(2021)
Article
Psychiatry
Hannah Marie Loso, Sarahjane Locke Dube, Bader Chaarani, Hugh Garavan, Matthew Albaugh, Masha Ivanova, Alexandra Potter
Summary: The study quantified sex differences in psychopathology among 9 and 10-year-olds, finding that males have higher scores and greater frequency of clinically meaningful levels of psychopathology. Puberty did not interact with sex to affect psychopathology, but as puberty advanced, the percentage of males and females with elevated scores increased.
PSYCHIATRY RESEARCH
(2021)
Article
Psychology, Clinical
Seon-Kyeong Jang, Gretchen Saunders, MengZhen Liu, Yu Jiang, Dajiang J. Liu, Scott Vrieze
Summary: This study uses genetically informative methods to investigate the genetic etiology of substance use and psychiatric disorders. The findings suggest that there is a high degree of pleiotropy and shared neurodevelopmental pathways, neurotransmission, and intracellular trafficking involved in these conditions.
PSYCHOLOGICAL MEDICINE
(2022)
Article
Ecology
Suchinta Arif, Aaron MacNeil
Summary: Ecologists often rely on observational data to understand causal relationships, but predictive techniques are not suitable for drawing causal conclusions. Instead, valid causal inference methods such as the backdoor criterion can be used to determine causal relationships in observational studies.
Article
Automation & Control Systems
Yuansi Chen, Peter Buehlmann
Summary: Domain adaptation is crucial in statistical machine learning when training and testing data differ. A theoretical framework based on structural causal models helps analyze and compare the performance of various domain adaptation methods, outlining necessary assumptions. The proposed CIRM method outperforms existing ones in handling perturbations in target data covariates and label distributions.
JOURNAL OF MACHINE LEARNING RESEARCH
(2021)
Article
Psychology, Developmental
Robert J. Loughnan, Clare E. Palmer, Carolina Makowski, Wesley K. Thompson, Deanna M. Barch, Terry L. Jernigan, Anders M. Dale, Chun Chieh Fan
Summary: Many behaviors show shared genetic influences among youth, but emerging psychopathologies also exhibit specificity for individuals with high genetic risk.
JOURNAL OF CHILD PSYCHOLOGY AND PSYCHIATRY
(2022)
Article
Psychology, Developmental
William E. Copeland, Sherika N. Hill, Lilly Shanahan
Summary: The objective of this study was to examine the associations between early cannabis use and adult functioning using four commonly used definitions of early cannabis use-related to timing, dose, duration, and associated symptoms. The results showed that daily, continued-over-time cannabis use beginning in adolescence had the most detrimental impact on a range of adult outcomes. Stopping early use did not completely eliminate later risks, but was associated with fewer negative outcomes.
JOURNAL OF THE AMERICAN ACADEMY OF CHILD AND ADOLESCENT PSYCHIATRY
(2022)
Article
Automation & Control Systems
Burak Varici, Karthikeyan Shanmugam, Prasanna Sattigeri, Ali Tajer
Summary: This paper investigates the problem of designing an optimal sequence of interventions in a causal graphical model to minimize cumulative regret. Two algorithms are proposed to address the issue. The key idea of these algorithms is to estimate the model parameters instead of directly estimating the reward distributions.
JOURNAL OF MACHINE LEARNING RESEARCH
(2023)
Article
Psychology, Clinical
Carter J. Funkhouser, Kelly A. Correa, Allison M. Letkiewicz, Eugene M. Cozza, Ryne Estabrook, Stewart A. Shankman
Summary: The study compared the reliability and validity of several models for different dimensions of psychopathology, finding concerns about the interpretability of psychopathology dimensions in bifactor models.
JOURNAL OF ABNORMAL PSYCHOLOGY
(2021)