arXiv subjects

Kang Shuai

Publications and source records attributed to Kang Shuai.

9 recordsLinked to original sources

arXiv2026

Identifiability of causal effects with non-Gaussianity and auxiliary covariates

Assessing causal effects in the presence of unmeasured confounding is challenging. Although auxiliary variables, such as instrumental variables, are commonly used to identify causal effects, they are often unavailable in practice due to stringent and untestable conditions. To address this issue, previous researches have utilized linear structural equation models to show that the causal effect is identifiable when noise variables of the treatment and outcome are both non-Gaussian. In this paper, we investigate the problem of identifying the causal effect using the auxiliary covariate and non-Gaussianity from the treatment. Our key idea is to characterize the impact of unmeasured confounders using an observed covariate, assuming they are all Gaussian. We demonstrate that the causal effect can be identified using a measured covariate, and then extend the identification results to the multi-treatment setting. We further develop a simple estimation procedure for estimating causal effects and derive a $\sqrt{n}$-consistent estimator. Finally, we evaluate the performance of our estimator through simulation studies and apply our method to investigate the effect of the trade on income.

stat.ME↗

arXiv2026

Identifying Causal Effects Using Instrumental Variables from the Auxiliary Dataset

Instrumental variable approaches have gained popularity for estimating causal effects in the presence of unmeasured confounders. However, the availability of instrumental variables in the primary dataset is often challenged due to stringent and untestable assumptions. This paper presents a novel method to identify and estimate causal effects by utilizing instrumental variables from the auxiliary dataset, incorporating a structural equation model, even in scenarios with nonlinear treatment effects. Our approach involves using two datasets: one called the primary dataset with joint observations of treatment and outcome, and another auxiliary dataset providing information about the instrument and treatment. Our strategy differs from most existing methods by not depending on the simultaneous measurements of instrument and outcome. The central idea for identifying causal effects is to establish a valid substitute through the auxiliary dataset, addressing unmeasured confounders. This is achieved by developing a control function and projecting it onto the function space spanned by the treatment variable. We then propose a three-step estimator for estimating causal effects and derive its asymptotic results. We illustrate the proposed estimator through simulation studies, and the results demonstrate favorable performance. We also conduct a real data analysis to evaluate the causal effect between vitamin D status and body mass index.

stat.ME↗

arXiv2026

Mediation analysis with unmeasured confounding between parallel mediators and outcome

Mediation analysis extending beyond single mediators has gained significant attention in recent years. However, related methods often assume the absence of unmeasured mediator-outcome confounding. To address this, we develop a mediation analysis framework that accounts for such confounding within a linear structural equation model with parallel mediators. Specifically, we introduce a pseudo proxy variable to capture unmeasured confounding, allowing us to identify causal parameters. Leveraging this proxy, we propose a partially penalized method to identify mediators that significantly affect the outcome. The resultant estimates are consistent, and the estimates of nonzero parameters are asymptotically normal. Motivated by these results, we further introduce a procedure that can consistently select active mediation pathways with large probability. Simulation studies demonstrate the superior performance of the proposed approach. Finally, we apply our approach to genomic data, identifying gene expressions that potentially mediate the impact of a genetic variant on mouse obesity.

stat.ME↗

arXiv2026

Bidirectional causal inference for binary outcomes in the presence of unmeasured confounding

Bidirectional causal relationships arising from mutual interactions between variables are commonly observed within biomedical, econometrical, and social science contexts. When such relationships are further complicated by unobserved factors, identifying causal effects in both directions becomes especially challenging. For continuous variables, methods that utilize two instrumental variables from both directions have been proposed to explore bidirectional causal effects in linear models. However, the existing techniques are not applicable when the key variables of interest are binary. To address these issues, we propose a structural equation modeling approach that links observed binary variables to continuous latent variables through a constrained mapping. We further establish identification results for bidirectional causal effects using a pair of instrumental variables. Additionally, we develop an estimation method for the corresponding causal parameters. We also conduct sensitivity analysis under scenarios where certain identification conditions are violated. Finally, we apply our approach to investigate the bidirectional causal relationship between heart disease and diabetes, demonstrating its practical utility in biomedical research.

stat.ME↗

arXiv2025

Episodic planetesimal disruptions triggered by dissipation of gas disk

Catastrophic disruptions of planetesimals occur in high-velocity collisions. Radioisotope dating of planetesimal disruption events recorded in meteorites confirms frequent catastrophic collisions in the first 10 Myr of the Solar System, reflecting a violent environment of the time. However, the nebula gas can damp the eccentricity of planetesimals and suppress the frequency of planetesimal collisions. Strong dynamical mechanisms that excited the protoplanetary disk are required. Here we show that the sweeping secular resonances of Jupiter and Saturn induced by the nebular gas dissipation, together with the mean motion resonances of Jupiter, can trigger a large number of catastrophic collisions, which occur episodically when the secular resonances are at 2-3 astronomical units and continue thereafter. After the gas dissipation completes, catastrophic collisions decrease in frequency, with scattering by planetary embryos becoming the major driving force of the collisions. Our results suggest that the violent environment excited by secular and mean motion resonances can be ubiquitous in protoplanetary disks during nebula dissipation.

astro-ph.EP↗

arXiv2025

Identification and estimation of causal peer effects using instrumental variables

In social science researches, causal inference regarding peer effects often faces significant challenges due to homophily bias and contextual confounding. For example, unmeasured health conditions (e.g., influenza) and psychological states (e.g., happiness, loneliness) can spread among closely connected individuals, such as couples or siblings. To address these issues, we define four effect estimands for dyadic data to characterize direct effects and spillover effects. We employ dual instrumental variables to achieve nonparametric identification of these causal estimands in the presence of unobserved confounding. We then derive the efficient influence functions for these estimands under the nonparametric model. Additionally, we develop a triply robust and locally efficient estimator that remains consistent even under partial misspecification of the observed data model. The proposed robust estimators can be easily adapted to flexible approaches such as machine learning estimation methods, provided that certain rate conditions are satisfied. Finally, we illustrate our approach through simulations and an empirical application evaluating the peer effects of retirement on fluid cognitive perception among couples.

stat.ME↗

arXiv2024

Spatial distribution of isotopes and compositional mixing in the inner protoplanetary disk

The mass-independent isotopic signatures of planetary bodies have been widely used to trace the mixing and transport processes in planet formation. The observed isotopic variations among meteorites have been further linked to the modeled mass-weighted mean initial semimajor axes, assuming a spatial isotopic gradient in the inner protoplanetary disk. However, nucleosynthetic isotopic anomalies of nonvolatile elements and mass-independent oxygen isotopic variation ($Δ^{17}$O) show different relationships with distance from the Sun. Therefore, it is crucial to know whether isotopes were distributed systematically with heliocentric distance. In this study, we performed N-body simulations on compositional mixing during the collisional accretion and migration of planetary bodies to investigate the spatial distributions of Cr and O isotopes in the inner protoplanetary disk. The modeled mass-weighted mean initial semimajor axes of the parent bodies of noncarbonaceous (NC) meteorites and terrestrial planets were used to calculate the isotopic compositions of these bodies. Our simulations successfully reproduced the observed nucleosynthetic Cr isotopic anomaly among Earth, Mars, and the NC meteorite parent bodies, consistent with a spatial gradient of isotopic anomalies in the inner disk. Asteroids originating from different regions in the inner disk were transported to the main belt in our simulations, resulting in the Cr isotopic anomaly variation of the NC meteorite parent bodies. However, the $Δ^{17}$O distribution among the terrestrial planets and the NC meteorite parent bodies could not be reproduced assuming a $Δ^{17}$O gradient. The absence of a $Δ^{17}$O gradient reflects that the oxygen isotopic mass-independent fractionation might have altered the spatial distribution of the nucleosynthetic $Δ^{17}$O variation before protoplanets formed.

astro-ph.EP↗

arXiv2024

Metal-silicate mixing in planetesimal collisions

Impacts between differentiated planetesimals are ubiquitous in protoplanetary discs and may mix materials from the core, mantle, and crust of planetesimals, thus forming stony-iron meteorites. The surface composition of the asteroid (16) Psyche represents a mixture of metal and non-metal components. However, the velocities, angles, and outcome regimes of impacts that mixed metal and silicate from different layers of planetesimals are debated. Our aim is to investigate the impacts between planetesimals that can mix large amounts of metal and silicate, and the mechanism of stony-iron meteorite formation. We used smooth particle hydrodynamics to simulate the impacts between differentiated planetesimals with various initial conditions that span different outcome regimes. In our simulations, the material strength was included and the effects of the states of planetesimal cores were studied. Using a statistical approach, we quantitatively analysed the distributions of metal and silicate after impacts. Our simulations modelled the mass, depth, and sources of the metal-silicate mixture in different impact conditions. Our results suggest that the molten cores in planetesimals could facilitate mixing of metal and silicate. Large amounts of the metal-silicate mixture could be produced by low-energy accretional impacts and high-energy erosive impacts in the largest impact remnant, and by hit-and-run and erosive impacts in the second-largest impact remnant. After impact, most of the metal-silicate mixture was buried at depth, consistent with the low cooling rates of stony-iron meteorites. Our results indicate that mesosiderites potentially formed in an erosive impact, while pallasites potentially formed in an accretional or hit-and-run impact. The mixing of metal and non-metal components on Psyche may also be the result of impacts.

astro-ph.EP↗

arXiv2024

Accretion regions of meteorite parent bodies inferred from a two-endmember isotopic mixing model

The diverse isotopic anomalies of meteorites demonstrate that the protoplanetary disk was composed of components from different stellar sources, which mixed in the disk and formed the planetary bodies. However, the origin of the accretion materials of different planetary bodies and the cosmochemical relationship between these bodies remain ambiguous. The noncarbonaceous (NC) planetary bodies originate from the inner solar system and have isotopic compositions distinct from those of the carbonaceous (CC) bodies. We combined Ca, Ti, Cr, Fe, Ni, Mo, and Ru isotopic anomalies to develop a quantitative two-endmember mixing model of the NC bodies. Correlations of the isotopic anomalies of different elements with different cosmochemical behaviors originate from the mixing of two common endmembers. Using this mixing model, we calculated the isotopic anomalies of NC bodies for all the considered isotopes, including the isotopic anomalies that are difficult to measure or have been altered by spallation processes. The mixing proportion between the two endmembers in each NC body has been calculated as a cosmochemical parameter, which represents the compositional relationship of the accretion materials between the NC bodies. Using the calculated mixing proportions, the feeding zones of the NC bodies could be estimated. The estimated feeding zones of NC bodies indicate a large population of interlopers in the main asteroid belt and an indigenous origin of Vesta. The feeding zones estimated in different planet formation scenarios indicate that the orbits of Jupiter and Saturn during formation of terrestrial planets were likely to be more circular than their current ones.

astro-ph.EP↗