Improved variance estimation in steady-state and time-dependent Monte Carlo neutron transport using history-based batch method in the iMC code

PGVAE-VBAKF: A robust strategy for complex system response prediction and noise variance estimation considering modeling errors and nonstationary noises

Intrinsic Minimum Average Variance Estimation for Dimension Reduction with Symmetric Positive Definite Matrices and Beyond

International Perspectives on the Covariation Among Adolescent Risk Behaviors.

Inverse probability of treatment weighting (IPTW) is a common approach to infer causal treatment effects when covariates are imbalanced at baseline or over time among treatment groups. One limitation of the IPTW is that a few observations with large weights can disproportionately influence inference, leading to dramatically increased variability in estimation. Stabilizing weights were developed to mitigate such a variability caused by excessively large weights. Since then, stabilizing weights have been widely regarded as good practice for IPTW, despite some misunderstandings and misinterpretations of their functionality. For example, a common misconception is that the original IPTW artificially inflates a study's statistical power because the weighted sample size appears to double that of the original. This article clarifies the role of stabilization in IPTW analysis, focusing on linear, logistic, and Cox's Proportional Hazard analyses in baseline binary treatment settings, which are commonly encountered in the regulatory space. Through theoretical derivations and simulation studies, we show that stabilized IPTW models yield identical point estimates to the original IPTW models in saturated linear and logistic regressions but yield slightly different estimates in Cox regressions. Stabilizing IPTW improves variance estimation over the original IPTW only when within-subject correlation due to weighting is ignored (as in model-based variance estimation, which is an incorrect approach), while none to minimal differences are observed when using robust variance estimation. Regardless of stabilization, a robust, sandwich-type variance estimator or resampling-based methods are the more appropriate approach for accurate variance estimation.

Modern techniques in variance estimation with auxiliary information: a logarithmic perspective

<table width="593" border="1" cellspacing="0" cellpadding="0"><tbody><tr><td valign="top" width="387"><p>Dengue Hemorrhagic Fever (DHF) exhibits substantial variation across districts and over time in Aceh Province, making it suitable for analysis within a panel data framework. This study models district-level DHF incidence using applied econometric techniques based on non-spatial panel data regression, employing a balanced panel dataset of 23 districts/cities observed from 2020 to 2022. The Common Effect Model (CEM), Fixed Effect Model (FEM), and Random Effect Model (REM) are estimated and formally compared using the Chow test, Hausman test, and Lagrange Multiplier test, with results consistently indicating that the Fixed Effect Model is the most appropriate specification due to the presence of unobserved, time-invariant district-specific effects. Diagnostic testing identifies heteroskedasticity in the error structure; therefore, the selected FEM is re-estimated using White cross-section robust standard errors to ensure reliable statistical inference. Empirical results show that population density is positively and statistically significantly associated with DHF cases, while the number of health workers is negatively and significantly associated, whereas rainfall, number of hospitals, sanitation coverage, and poverty level do not exhibit statistically significant effects in the final robust specification. The selected model explains approximately 86% of the within-district variation in DHF incidence, demonstrating the importance of appropriate model specification and robust variance estimation in panel data regression applied to epidemiological outcomes, while emphasizing that the estimated relationships represent statistical associations rather than causal effects.</p></td></tr></tbody></table>

Combining multiple biomarkers into a single diagnostic score can improve disease classification. However, traditional methods such as logistic regression and linear discriminant analysis depend on restrictive distributional assumptions, which can limit their effectiveness when dealing with complex or heterogeneous datasets. To address these limitations, more practical methodological alternatives are required. This study investigates the utility of two sufficient dimension reduction (SDR) methods, Minimum Average Variance Estimation (MAVE) and the Outer Product of Gradients (OPG), for constructing composite biomarker scores in binary diagnostic classification. A comprehensive simulation study was conducted under four data-generating scenarios, varying in sample sizes, mean shifts, variance heterogeneity, normal and non-normal distributional forms. SDR based scores were constructed using likelihood ratio statistics under both the central subspace and central mean subspace frameworks. Classification performance was evaluated using the area under the receiver operating characteristic curve (AUC). Traditional methods, logistic regression and linear discriminant analysis, were included as benchmarks. To demonstrate practical utility, the methods were applied to the Breast Cancer Wisconsin (Diagnostic) dataset. In simulations, SDR methods outperformed traditional approaches consistently in settings with variance heterogeneity and mixture structures, yielding higher AUC values. The performance of SDR methods was less robust where the data had strong skewness, though it remained comparable to that of traditional methods. In the real dataset, SDR methods achieved similar discriminative accuracy to traditional methods while offering more compact and interpretable summaries of biomarker contributions. SDR methods combined with likelihood-based scoring show potential for a relatively robust and interpretable framework in the context of this paper. They are particularly advantageous in settings with complex diagnostic structures, while maintaining competitiveness in well-structured data. These findings support the use of SDR methods as practical tools for combining biomarkers in precision medicine.

Sexual minority youth report significantly higher rates of substance use than heterosexual youth, yet a comprehensive and systematic evaluation and synthesis of the magnitude of this inequity has not been conducted. The purpose of this study was to conduct a meta-analysis to assess the magnitude of overall inequities in substance use between sexual minority and heterosexual youth and examine demographic and methodological moderating factors that might impact variability in substance use patterns. A meta-analysis of studies that examined sexual orientation inequities in substance use. We conducted a comprehensive literature search across four electronic databases (PubMed, APA PsycINFO, Web of Science, ProQuest; January 2008-July 2024). Studies were eligible if they included a youth participant population (mean sample age of 25 years or younger), examined differences in substance use between sexual minority and heterosexual groups, and published between 2008 and 2024 in English. Primary outcomes were any measures of substance use; secondary outcome was age of substance use initiation. Following PRISMA guidelines, reviewers independently reviewed and extracted data. Analyses employed random-effects models, with robust variance estimation to account for dependency among multiple effect sizes within studies. Analyses examined continuous and dichotomous outcomes separately. Bivariate meta-regression examined moderators. Among 304 studies of 5 928 282 youth, sexual minority youth reported greater quantity and frequency of all assessed substances (i.e. alcohol, nicotine, cannabis, prescription drugs, powder cocaine, crack cocaine, meth/amphetamine, 3,4-methylenedioxymethamphetamine, heroin, and other substances) and engaged in more polysubstance use than heterosexual youth. For continuous outcomes, Hedges' g ranged from 0.10 (95% confidence interval [CI], 0.05-0.15, I2= 99.52, participants = 354 201, studies = 48) for alcohol to 0.40 (95% CI, 0.21-0.59, I2= 100.00, participants = 30 679, studies = 5) for mixed/polysubstance use. Dichotomous outcomes showed consistently elevated odds ratios, ranging from 1.34 (95% CI: 1.24-1.46, I2= 96.01, participants = 3 500 203, studies = 128) for alcohol to 4.63 (95% CI, 2.91-7.38, I2= 86.35, participants = 391 827, studies = 10) for heroin use. Sexual minority youth also had earlier ages of initiation (all substance outcomes: odds ratio, 1.45; 95% CI, 1.04-2.03, I2= 95.39, participants = 619 187, studies = 11). Moderation results indicated that inequities were larger for plurisexual youth (e.g. bisexual, pansexual), sexual minority girls and young women, and adolescents 18 years of age or younger. The magnitude of inequities was also larger for lifetime measures of use compared with measures of recent use. Sexual minority youth - particularly those who are plurisexual, sexual minority girls and young women, and adolescents 18 years of age or younger - appear to engage in substance use at higher rates than heterosexual youth.

Abstract This meta-analytic work, conducted in accordance with PRISMA guidelines, evaluates the efficacy of digital interventions in improving reading outcomes in children under 15 years of age. We included 41 studies ( k = 194) focused on poor readers and 15 studies ( k = 69) on general school readers, all employing randomized controlled or quasi-experimental designs published between 2000 and 2022. Multilevel meta-analyses with robust variance estimation, conducted using the open-access tool ShinyApp, revealed a medium effect for poor readers ( g = 0.433, CI = [0.3, 0.56], p < .001), corresponding to substantial literacy gains, and a smaller effect for general readers ( g = 0.256, CI = [0.08, 0.43], p = .009) reflecting more modest progress. Notably, interventions targeting reading-specific skills, such as decoding and comprehension, were effective across both groups, whereas those focused on domain-general cognitive skills, such as attention and working memory, benefited only poor readers. Results varied slightly by orthographic transparency, with numerically stronger effects in languages with simpler syllabic structures. No significant differences emerged between decoding and comprehension outcomes. These findings highlight the potential of tailored digital interventions to enhance reading—especially for children facing literacy challenges— while underscoring the need to adapt them to diverse learner and contextual characteristics to ensure scalability and cross-linguistic applicability.

Malnutrition and food insecurity remain major public health issues in Ethiopia. Although several studies in the Sidama region have explored their association, few have jointly analyzed these outcomes using an integrated statistical framework. Identifying shared and unique predictors is crucial for developing context-specific prevention strategies. A cross-sectional study was conducted from November 2023 to February 18, 2024, involving 1,149 households with children under five years of age. Participants were selected using a multistage sampling technique. Data were collected through a pre-tested structured questionnaire administered by trained data collectors. Child nutritional status was assessed using height-for-age Z-scores (HAZ), where children with HAZ < - 2 SD were coded as stunted (1), and those with HAZ ≥ - 2 SD were coded as not stunted (0). The Household Food Insecurity Access Scale (HFIAS) was used to categorize households according to food access levels using its nine standard questions. Joint generalized linear mixed models were applied to identify predictors of both stunting and food insecurity and assess the correlation between them. Statistical significance was set at p < 0.05. Out of the total households surveyed with under-five children, 801 (78.2%; 95% CI: 75.2-83.7) were found to be food insecure, and 541 children (47.08%; 95% CI: 42.5-51.5) were stunted. The random effects in the joint generalized linear mixed model indicated significant variability across clusters (Kebeles), with variance estimates of 0.30 (p = 0.001) for food insecurity and 0.45 (p < 0.001) for stunting. A moderate positive correlation of 0.52 (p = 0.039) was observed between the two outcomes. Significant predictors for both stunting and food insecurity included: employment status, father's age, number of under-five children in the household, mother's age at first birth, succeeding birth interval, household wealth index, husband's occupation, parental education levels, dietary diversity score, and meal frequency per day. This study revealed a high prevalence of stunting and food insecurity among households with under-five children in Hawassa Zuria district, Sidama region, Ethiopia. Household and child-level factors were significantly associated with these conditions. Integrated, multisectoral interventions should prioritize vulnerable households, focusing on dietary diversity, maternal and child health services, and sustainable food access. Policymakers should strengthen community-based programs to tackle both the immediate and underlying causes of child malnutrition and household food insecurity.

Background/Objectives: Special diets can be required for medical, religious, cultural, or ethical purposes. This study examined the relationship between the organization of school nutrition and the availability of special diet meals among students in public primary and secondary schools in Kraków (Poland). Methods: An observational study was conducted in 2022 using a web-based survey targeting managers of primary (n = 68) and secondary schools (n = 18), as well as parents of attending students (n = 1730). Factors associated with providing special diets were analyzed using generalized linear models with robust variance estimators. Results: According to school managers, the availability of special diet meals was associated with employing a dietitian responsible for menu planning, the presence of students with disability certificates, students’ participation in school meal programs, and higher per-child nutrition costs. Based on parental reports, 16.01% of all students followed a special diet, most often due to medical recommendations, with a higher prevalence observed among secondary school students (26.7%). Special diets were reported more frequently for children with food intolerances and allergies, obesity, chronic conditions, or disability certificates. Adjusted models also indicated slightly higher probabilities of being on a special diet among students attending secondary schools or sports classes compared with their peers. Conclusions: Improving the availability of special diet meals in schools requires legislative action, adequate funding, and institutional support, including investments in kitchen infrastructure and the employment of dietitians. These measures are particularly important in institutions enrolling children with disabilities to ensure equitable access to appropriate nutrition.

Survey sampling is a widely used technique for collecting data from a subset of a bigger population. Among its methods, stratified random sampling is particularly valuable for yielding precise inferences about distinct subgroups within a population by dividing the population into mutually exclusive strata and sampling from each group. This approach reduces sampling error and enhances the accuracy of population estimates. In this study, we propose a set of improved calibrated log-ratio-type estimators for estimating population variance under a stratified sampling framework. The performance of three proposed estimators is evaluated and compared in terms of the mean squared error. A simulation study is conducted to assess the efficiency of the estimators, complemented by a real-life application to validate the simulation results. The findings demonstrate that the proposed calibrated log-ratio variance estimators outperform existing methods by achieving lower mean squared error.

ABSTRACT Working with an Indigenous Advisory Committee, including an Inuit Health Advisor and Researcher, we analyzed the 2017 Aboriginal Peoples Survey to examine prevalence and factors associated with pain-related disabilities (PRDs) among Inuit in Canada. Self-reported data were collected from Inuit ≥15 years. PRDs were defined as ‘sometimes’, ‘often’, or ‘always’ experiencing activity limitations due to pain from a long-term condition lasting ≥ six months. We computed PRD prevalence [95% CI] overall, and by geographic location, age, sex, type and number of co-existing disabilities. Modified Poisson regression with robust variance estimation modelled associations between Inuit social determinants of health and PRDs. Person-level and bootstrap weights were applied for all analyses. Among Inuit, 11.1% [10.0, 12.4] reported PRDs. Females [13.4% (11.8, 15.1)], individuals 55 + [23.7% (21.6, 25.9)], and those who lived outside Inuit Nunangat [17.1% (14.1, 20.5)] experienced higher prevalence of PRDs. Prevalence increased with the number of disabilities—highest among those with co-existing physical disabilities. Additionally, higher education, residential school attendance, and those who experienced difficulties related to food, housing, employment, and health were more likely to report PRDs. Characteristics which may increase the risk of PRDs need to be shared with Inuit stakeholders to guide next steps for awareness, advocacy, services and interventions.

ABSTRACTAlthough count data are collected in many experiments, their analysis remains challenging, especially in small sample sizes. Until now, linear or generalized linear models in Poisson or Negative Binomial distributional families have often been used. However, these data frequently show signs of over‐, underdispersion, or even zero‐inflation, casting doubt on these distributional assumptions and leading to inaccurate test results. Since their distributions are usually skewed, data transformations (e.g., log‐transformation) are not unusual. This underscores the need for statistical methods not to hinge on specific distributional assumptions. We delve into multiple contrast tests that allow general contrasts (e.g., many‐to‐one or all‐pairs comparisons) to analyze count data in multi‐arm trials. The methods vary in their effect and variance estimation, as well as in approximating the joint distribution of multiple test statistics, including frequently used methods such as linear and generalized linear models, and data transformations. An extensive simulation study demonstrates that a resampling version effectively controls the Type I error rate in various situations, while also highlighting the method's limitations, including overly liberal Type I error rates. Some standard methods, which have inflated Type I error rates, further underscore the need for alternative approaches. Real data applications further emphasize the applicability of these methods.

Comparing multiple people each to the grand mean of log-normally distributed endpoints.

The merit of superimposed vibration for flexibility and passive stiffness: A systematic review with multilevel meta-analysis.

Posttraumatic stress disorder symptoms and suicide ideation, attempt, and risk among active-duty service members and veterans: A systematic review with three meta-analyses of associations and moderators.

Log-transformed approaches to variance estimation using auxiliary data

Seroepidemiology of hepatitis C virus infection in people aged 1-69 years in China: a national, cross-sectional study.