Many processes related to status, power, and influence within social networks have been modeled using forced linear diffusion models; examples include the highly successful Friedkin-Johnsen model of social influence, the status/power scores of Katz and Bonacich, and the widely used network autocorrelation model. While a basic assumption of such models is that the impact of one individual on another through any given path falls exponentially with path length, the total impact of the first individual on the second involves contributions from walks of all lengths; thus, while total impact is expected to decline with network distance, the relationship is not trivial. Here, we provide an approximate solution for the total impact of one node on another as a function of network distance, showing that the total impact is given to first order by a product of eigenvector centrality scores together with an expression in terms of the graph spectrum (eigenvalues of the adjacency matrix) that falls exponentially with distance. We also show how this solution can be refined using higher-order eigenvectors of the adjacency matrix. A numerical study on interpersonal networks drawn from educational settings verifies an average exponential decline in impact strength under the linear diffusion model, and shows that the first-order eigenvector approximation can often be a good proxy for total impact as obtained from the exact solution. This suggests a simple model that can be used to approximate total impact for social influence or status processes in a range of settings.
Ultrasonic infrared thermography (SonicIR) is a powerful technique for nondestructive defect detection. Existing deep-learning methods typically rely on spatial thermal images or temporal thermal signals alone, while recent fusion approaches mainly employ feature concatenation, element-wise aggregation, or unidirectional interactions that inadequately model the intrinsic correlations between heterogeneous spatiotemporal signals. Consequently, the complementary and difference information between spatial and temporal representations remains insufficiently exploited, limiting defect discrimination under complex industrial interference conditions. To address these challenges, this paper proposes a Dual-branch Feature Fusion Classification Network (DFFC-Net) for automated ROI-based SonicIR defect detection. The framework jointly utilizes Maximum-Normalized Mean (MT) and Correlation-Time (CT) signals to characterize both thermal diffusion dynamics and spatial evolution. Unlike conventional fusion methods that merely aggregate features, the proposed Cross-Information Fusion Module (CIFM) explicitly decomposes and integrates shared correlation information and complementary difference information through a cross-attention mechanism. Furthermore, a Bidirectional Feature Fusion Module (BFFM) enables mutual information exchange between heterogeneous signal streams, allowing pixel-level spatiotemporal association modeling rather than one-way feature injection. To further enhance robustness, an Enhanced Spatial and Channel Attention Module (ESCM) jointly explores channel-wise dependencies and spatial contextual relationships, improving the discrimination of defect-sensitive features under interference. Experiments conducted on a dataset comprising 153 SonicIR sequences of Titanium and Inconel alloy specimens demonstrate that the proposed method consistently outperforms state-of-the-art approaches in defect detection accuracy. The results confirm that explicitly modeling both shared and complementary spatiotemporal information provides a more effective solution for SonicIR defect identification than conventional feature fusion strategies.
Background Few evidence-based treatments exist for addressing pain among patients receiving medication for opioid use disorder (MOUD). To inform novel interventions, this exploratory study investigated pain profiles and their associations with multi-level individual, family, neighborhood, social and community correlates among patients receiving MOUD. Methods Using baseline data from the NIH HEAL-funded longitudinal Collaboration Linking Opioid Use Disorder and Sleep (“CLOUDS”) study to examine sleep and circadian mechanisms that contribute to nonmedical opioid use and MOUD retention, we investigated pain profiles using k-means cluster analysis, and compared individual, family, neighborhood, social and community correlates using generalized linear models. Results The sample included 211 participants receiving MOUD (mean age 41.4 years; 42.2% female). Four pain profiles emerged. Profile A (28.4%): low pain, low catastrophizing, and high resilience. Profile B (28.4%): low pain, low catastrophizing, and low resilience. Profile C (23.2%): moderate pain, moderate catastrophizing, and high resilience. Profile D (20%): moderate pain, moderate catastrophizing, and low resilience. Varying individual, family, neighborhood, and social and community correlates were identified across profiles. Generally, participants in profile A had the lowest psychiatric distress and the highest physical and psychological health, whereas the converse was true for participants in profile D. Conclusion Patients receiving MOUD in profile D had a less favorable pattern of individual, family, neighborhood, social and community correlates. Further research is needed to confirm these findings as they may eventually guide tailored multi-level integrated treatments for patients receiving MOUD, depending on their pain profile.
Abstract The US construction industry grapples with acute labor shortages and escalating wage pressures driven by the surge of megaprojects such as data centers, semiconductor fabrication facilities, and large-scale transportation infrastructure. These wage pressures do not arise uniformly across the industry but often emerge within tightly interconnected skilled trade labor markets, where wage changes in one trade may propagate to others through labor mobility, competitive hiring, and coordinated project delivery requirements. However, existing construction labor research has largely treated construction trade wages as independent processes, overlooking the potential for wage transmission and spillover effects across functionally linked trades. This study addresses this gap by examining wage transmission mechanisms and measuring wage spillover effects between two major construction trades, electrical and mechanical contractors. Quarterly wage data for electrical and mechanical construction contractors across 12 US Midwestern states and at the national level from 2000 to 2024 were collected from the US Census Bureau and analyzed using multivariate time-series methods. The novel contributions of this research include: (1) examining wage transmission and spillover effects between two functionally interdependent trades (i.e., electrical and mechanical contractors); (2) identifying wage interdependencies at both the US national level and across 12 Midwest states, capturing regional heterogeneity and localized labor market dynamics; (3) uncovering asymmetric lead-lag wage relationships between electrical and mechanical contractors; and (4) quantifying the dynamic propagation of wage shocks, providing predictive insights that static, cross-sectional analyses cannot capture. The findings of this research are expected to assist contractors, owners, policymakers, and other construction stakeholders in understanding dynamic wage transmissions and spillover effects, enabling more accurate labor cost estimation, improved contract escalation provisions, and regionally targeted workforce development strategies.
Urban flooding alters travel behavior in coastal communities, yet few studies jointly examine multiple trip purposes, warning channels, and normal–flood mode choice. This study analyzes travel responses to flood scenarios using resident surveys, discrete choice modeling, and elasticity analysis. Overall trip rates declined, with discretionary and educational travel showing the steepest reductions, while essential trips such as work and shopping were largely retained but often delayed. Flooding increases travel times across all trip purposes, with smaller delays for work trips and moderate disruptions for shopping and non-home-based discretionary travel. Mode choice behavior shows reduced driving alone and greater use of carpooling and public transportation. Under normal conditions, travel is shaped by income, vehicle ownership, and employment, but during flooding, flood warnings and temporal flexibility become dominant. Interaction and elasticity results show that information access and adaptability drive resilient mobility decisions, informing adaptive transportation planning.
