Boyoung Kim

This study analyzed how the perceived organizational silence climate held by members of digital platform service organizations is related to turnover intention through learned helplessness, psychological withdrawal, and job burnout. The perceived organizational silence climate is defined here as an individual-level psychological climate perception: the degree to which an individual employee perceives that a belief is prevalent within the organization that raising problems or expressing opinions is of no use or may lead to disadvantages. A cross-sectional online survey was conducted with 346 employees working at digital platform service firms in South Korea. After the validity of the measurement model was examined through confirmatory factor analysis, the relationships among the variables were analyzed using latent-variable structural equation modeling, and the model-implied indirect associations were estimated with bias-corrected bootstrap confidence intervals based on 5,000 resamples. The results showed that the perceived organizational silence climate was positively associated with learned helplessness, psychological withdrawal, and job burnout. All three variables were also positively associated with turnover intention. Each model-implied indirect association through learned helplessness, psychological withdrawal, and job burnout was statistically significant, whereas the direct association was not. Pairwise comparisons of the three specific indirect associations were not statistically significant, indicating that the three pathways could not be distinguished in magnitude. These findings extend the focus of organizational silence research from individual silence behavior to employees’ perception of the organizational silence climate and test, within a single parallel mediation model, the cognitive, behavioral-relational, and affective pathways that connect the organizational silence climate to turnover intention. The findings further suggest that, to manage turnover intention, it is necessary to diagnose employees’ perceptions of both the futility and the riskiness of voice and to improve organizational environments in which the usefulness of expressing opinions is perceived to be low or the possibility of disadvantage is perceived to be high. The data are cross-sectional and were collected from a single source. The estimates reported here therefore describe associations rather than causal effects, and the model-implied indirect associations require confirmation in longitudinal or multiwave designs.

Generative AI is increasingly embedded in knowledge work, yet employees may interpret the same technology either as a resource that extends their capabilities or as a threat involving replacement or monitoring. The psychological pathways linking this interpretation to job engagement and job embeddedness remain underexplored. Drawing on job demands–resources theory, this study positions perceived generative AI complementarity as a technology-based job resource and examines whether it is associated with job engagement and embeddedness through psychological empowerment and psychological safety. Data were collected through a cross-sectional online survey of 428 employees of digital platform service organizations in South Korea. A latent-variable structural equation model was used to evaluate the measurement model and all hypothesized associations, and path-specific indirect associations were tested using bias-corrected bootstrap resampling. Perceived generative AI complementarity was positively associated with both psychological resources, and both resources were positively associated with job engagement and job embeddedness. All four indirect associations were supported, whereas neither residual direct association was distinguishable from zero once measurement error was modeled; the two pathways did not differ significantly in magnitude. Because the data are cross-sectional, single-source, and self-reported, the findings describe associations consistent with the proposed model rather than causal effects.

This study examines the influence of consumers’ brand-related perceptions on brand fandom behaviors. In particular, the study focuses on how brand experience, consumers’ perceived product value, and perceived brand innovativeness affect sharing and advocacy behaviors as well as creation behaviors, mediated by self-transcendent emotion and self-brand connection. This study also investigated the moderating role of the need for self-expression in these mediating pathways. To test the proposed hypotheses in the context of high-tech brands, structural equation modeling was employed, and data were collected from 635 Tesla owners and prospective buyers in the United States and Republic of Korea. All ten hypothesized paths were significant. The two mediators, however, correlated only weakly with each other while each correlated strongly with the behaviors, indicating two largely independent routes to the same fan behaviors rather than two indicators of a single bond. The need for self-expression strengthened all four second-stage paths, so brand fandom is best described as a two-stage process with two distinct routes and a shared expressive condition. The findings suggest that designing for only one of the two routes reaches only part of a brand’s potential fan base.

Advancements in next-generation wireless communication technologies, such as 5G, have enabled the utilization of Ultra-Reliable Low-Latency Communication (URLLC) and Enhanced Mobile Broadband (eMBB) resources. To enhance user-perceived performance, or Quality of Experience (QoE), it is essential to appropriately allocate network resources to applications based on their specific requirements. Consequently, there is a growing necessity for intelligent resource management at the User Equipment (UE) level. This study proposes a three-stage on-device ML pipeline for service-aware power optimization in 5G UE. The pipeline consists of: (1) a Traffic Classifier (XGBoost) that performs real-time binary classification of RT/NRT traffic; (2) a NO-DATA Predictor (Student-LogReg with Knowledge Distillation) that forecasts traffic-idle intervals in NRT sessions; and (3) a Finite State Machine (FSM) Controller that selectively suspends classifier inference during predicted NO-DATA periods. The FSM Controller suspends classifier operation at P(NO-DATA) ≥ 0.80 for two consecutive slots and resumes at P(NO-DATA) < 0.60, enabling recovery within one slot. Experimental evaluations on real-world KT 5G SA network traffic data demonstrate that the proposed framework achieves a 61.42% reduction in classifier inference calls while maintaining a QoE error rate of 0.26% and zero RECLASSIFY events, confirming that the Predictor's DATA Recall of 98.35% effectively serves as a safety net against classifier misclassification. This approach provides a practical solution for implementing application-aware, energy-efficient network management in resource-constrained on-device environments. To validate generalizability beyond simulation, the pipeline was installed on a custom-built Android measurement application and field-tested on seven Samsung Galaxy device models across six countries (Korea, Japan, Australia, Mexico, Brazil, and the United Kingdom) on commercial networks, yielding 28 measurement sessions with RECLASSIFY = 0 in all cases. Real-world field measurements show reduction rates ranging from 19.6% to 60.4%, with NRT applications (YouTube, Netflix) achieving a mean of 46.1% compared to 33.2% for RT applications (YouTube Live, Zoom), confirming a 12.9 pp service-type gap consistent with the simulation results.

This study aims to empirically assess the effect of ethical leadership in corporate tax leaders on corporate transparency, ESG management, and corporate performance. The research model was constructed, and hypotheses were formulated for structural equation analysis. This study used a total of 262 questionnaires collected from tax leaders in large South Korean companies. A research model was designed using structural equations, and results were derived through confirmatory factor analysis, correlation analysis, and path analysis. The analysis results verified that ethical leadership had a positive effect on corporate transparency and ESG management. However, ethical leadership in tax leaders had no direct impact on corporate performance. This study also proved that this leadership can be positively influenced by corporate transparency and ESG management activities and is not a direct determinant of corporate performance. Additionally, corporate transparency had no direct effect on corporate performance; however, it can be affected along with ESG management as a mediated factor through which the ethical leadership of tax leaders influences corporate performance. These results indicate that the implication is that ethical leadership in corporate tax management impacts corporate performance when corporate transparency and ESG management activities are reinforced, and its effectiveness is maximized as well. Further, it suggests that corporate tax leaders have a crucial role in maintaining financial transparency, fair tax management, and social responsibility.

Follower recognition of leaders significantly influences organizational trust, role-model perception, and the intention to recommend leaders. While positive recognition can be shaped by leaders’ behaviors and competencies, their psychological resource, namely mental toughness, may also affect followers’ perceptions and evaluations. The purpose of this study was to empirically examine the effects of leaders’ mental toughness and leadership competence on follower recognition within organizational contexts. Using a large-scale 360-degree feedback dataset, we analyzed responses from approximately 1,500 employees evaluating 197 senior executives in South Korean corporations. The results showed that leadership competence had a significant positive effect on follower recognition. In contrast, mental toughness did not have a direct effect on recognition but exerted a full mediating effect through leadership competence. This suggests that mental toughness is not directly perceived by followers; rather, it becomes visible when translated into leadership behaviors, which subsequently lead to positive recognition. The study contributes theoretically and practically by reframing mental toughness as a core psychological resource operating through leadership competence, rather than a mere personal trait, in shaping leader–follower dynamics.

As high-tech companies become increasingly important globally, these companies dedicate greater resources to robust technology security and leakage-prevention measures. This study aims to define the factors influencing autonomy, competence, and relatedness which are the basic psychological needs that influence the information leakage-prevention behaviors of employees in high-tech industry enterprises based on self-determination theory (SDT). Additionally, this empirical study investigates the influence of these factors on the behavioral intentions related to industrial technology leakage prevention. It analyzes the relationship between basic psychological needs and intrinsic and extrinsic motivation based on SDT. Moreover, it verifies the mediating effect of motivation on the relationship between basic psychological needs and behavioral intentions. An online survey was conducted among employees of high-tech industries in Korea resulting in the collection of 200 questionnaires. The analysis demonstrated that autonomy, competence and relatedness had a significant positive effect on intrinsic motivation. It means that intrinsic motivation plays a more critical role than extrinsic motivation in shaping the behavioral intention of high-tech firms to prevent industrial-technology leakage. However, no effect was observed for relatedness on extrinsic motivation or for extrinsic motivation on behavioral intentions. The findings indicate that autonomy negatively impacted extrinsic motivation. This suggests that in the high-tech technology sector, employees’ intrinsic motivation primarily drives their technology security behaviors based on personal professional judgment and ethics rather than extrinsic motivation.

The popularization of generative AI has led to significant social and industrial changes globally. As generative AI technology rapidly evolves, its influence is expected to grow, increasing the need for research on its acceptance and use. This study empirically analyzes the relationship between user attitudes and the adoption intent of generative AI services, offering insights into their utilization. Based on the Unified Theory of Acceptance and Use of Technology (UTAUT), four factors—Performance Expectancy, Effort Expectancy, Facilitating Conditions, and Hedonic Motivation—were identified as components of adoption intent. Additionally, this study analyzed the causal relationship between these factors and user attitudes, mediated by users' perceived emotional and functional values. A structural equation model was constructed with data from 356 users of generative AI services in South Korea. The analysis revealed that Performance Expectancy and Facilitating Conditions influence user attitudes through emotional and functional value mediation. Effort Expectancy significantly affected functional value, while Hedonic Motivation influenced emotional value, both exhibiting mediating effects. Emotional value had a greater impact on user attitudes than functional value. These findings suggest that emotional experiences are critical in the adoption of generative AI services, highlighting the need for strategies to enhance user engagement and satisfaction.