Working from anywhere: yin-yang cognition paradoxes of knowledge sharing and hiding for developing careers in China

Rapid advances in ICT and COVID-19 quarantines have driven widespread adoption of working from anywhere (WFA), blending remote and on-site work. While WFA is increasingly common and may reshape careers, prior findings on its career impact are mixed. WFA may enhance autonomy and intrinsic motivation, improving career prospects, but may also reduce organizational monitoring and social ties, potentially harming outcomes. The Microsoft WFA experience highlights paradoxes in knowledge exchange: technology enables parallel knowledge sharing yet reduces synchronous communication and social connection, impeding knowledge transfer. This study adopts a yin-yang harmonizing cognitive frame to examine paradoxical, interdependent elements of knowledge sharing (KS) and knowledge hiding (KH) within WFA and their implications for career development (CD). Research questions: (1) How is WFA (operationalized as remote work time, RWT) related to CD? (2) How does the interaction of KS and KH affect the RWT–CD relationship? The study proposes: H1: an inverted U-shaped RWT–CD relationship; H2: KS×KH relates to CD; H3: the RWT–CD relationship is moderated by KS×KH; H4: the moderation is strongest when KS is high and KH is low. The work contributes by applying a yin-yang paradox lens to HRM and KM in WFA contexts and offers practical guidance for managing careers in volatile, digitized environments.

Theoretical foundation centers on yin-yang harmonizing, a paradoxical cognitive framework analogous to Western paradox theory, emphasizing dynamic, complementary opposites. Applied to WFA, RWT embodies opposing forces: autonomy/flexibility (motivating) versus isolation/monitoring loss (demotivating), suggesting a curvilinear link to CD. In knowledge exchange, KS (enabling mutual knowledge availability) and KH (intentional concealment) are interdependent rather than simple opposites. Tacit–explicit knowledge dialectics underscore that knowledge flow in WFA is complex: ICT facilitates parallel sharing but reduces embodied, synchronous communication, potentially increasing KH. Prior work associates KS with positive outcomes (commitment, innovation) and KH with strategic withholding influenced by motivation and justice. The study frames KS and KH as yin-yang elements whose interaction should shape CD, especially under varying RWT. Hypotheses posit an inverted U-shaped RWT–CD link (H1), a significant KS×KH association with CD (H2), moderation of the RWT–CD curve by KS×KH (H3), and strongest positive moderation when KS is high and KH low (H4).

Design: Quantitative, time-lagged survey with two waves over three months (initiated late April 2021) to reduce common method variance. Context and sampling: Seven manufacturing firms in Zhejiang Province, China, using hybrid remote/on-site work. HR contacts distributed online questionnaires via Sojump (wjx.cn) and WeChat using simple random sampling. Only full-time employees participated. Pre-study interviews with HR professionals and senior employees ensured measurement clarity and model logic. Participants: 364 usable matched responses (response rate 70.3%). Demographics: 181 women, 183 men; mean age 33.5 years; 64.6% married; 19.2% state-owned firms; firm sizes varied. Measures: 6-point Likert scales (strongly disagree to extremely agree) to mitigate midpoint bias. - Remote Work Time (RWT): self-reported proportion of average workweek teleworking (0–100%), mean 44%. - Knowledge Sharing (KS): peer-rated, adapted 11-item scale (Hu et al., 2009); example items include willingness to help and share knowledge; Cronbach’s α = 0.942. - Knowledge Hiding (KH): peer-rated, adapted from Connelly et al. (2012); example items include pretending not to know or refusing to answer; Cronbach’s α = 0.935. - Career Development (CD): proxied via career control (Akkermans et al., 2013) 4-item scale; example item: goal setting for career; Cronbach’s α = 0.911. Control variables: age, gender (0 male, 1 female), marital status (0 not married, 1 married), firm type (0 state-owned, 1 other), firm size (1 <80; 2: 81–500; 3: 501–1000; 4: 1001–2000; 5: >2000). Procedure and analysis: Wave 1 measured RWT, KS, KH; Wave 2 (three months later) measured CD. Reliability: All α > 0.70; CR > 0.7 and AVE > 0.5. CFA established discriminant validity: full model fit χ²(71)=237.337, χ²/df=3.342, CFI=0.960, TLI=0.948, RMSEA=0.080, SRMR=0.045; alternative models fit worse. CMV check via CFA marker method indicated no significant CMV. Hypothesis testing: Moderated hierarchical regression with mean-centering; sequentially added main effects, quadratic term (RWT²), two-way interactions (KS, KH, KS×KH), and three-way interactions involving RWT and RWT² with KS and KH. Model R² improved across steps (up to ~0.43). Interaction plots followed Dawson (2014).

• RWT–CD relationship is inverted U-shaped: RWT positively related to CD (β=1.918, p<0.001) and RWT² negatively related (β=−5.599, p<0.001), supporting H1. - Main effects: KS positively related to CD (β=0.647, p<0.01); KH negatively related to CD (β=−0.475, p<0.05). - KS×KH interaction positively related to CD (β=0.080, p<0.05), supporting H2. - Three-way moderation: The curvature of the RWT–CD relationship was significantly moderated by the joint effect of KS and KH; RWT²×KS×KH was significant (β≈−1.421, significant), supporting H3. - Conditional effects: RWT had the strongest positive impact on CD when KS was high and KH was low (supporting H4). - Measurement model fit was acceptable (CFA: CFI=0.960, TLI=0.948, RMSEA=0.080, SRMR=0.045); CMV unlikely. - Descriptives: Mean RWT 44%. - Controls: Positive associations with age (older), gender (female), and state-owned firm type; negative association with being unmarried and CD.

Findings address the research questions by demonstrating a nonlinear (inverted U-shaped) link between RWT and CD and by showing that knowledge dynamics (KS and KH) jointly shape career outcomes in WFA. Moderate levels of remote work optimize CD, consistent with a balance between autonomy and connection; too little or too much remote work can hinder development due to disruption or remote work fatigue/isolation. KS enhances, and KH diminishes, CD, but their interaction suggests they are interdependent yin-yang elements rather than simple opposites. The moderating role of KS×KH indicates that cultivating high sharing and minimizing hiding strengthens the benefits of remote work for career growth, especially at moderate RWT. The study contributes theoretically by applying a yin-yang paradox lens to HRM and KM in WFA, enriching understanding of flexible work arrangements and micro-level paradoxes during and after COVID-19. Practically, it suggests designing hybrid policies and knowledge practices that optimize career development by fostering trust, cohesion, and embodied communication in digital contexts.

In ICT-enabled WFA contexts that transcend traditional spatiotemporal work boundaries, knowledge exchange occurs across sophisticated, evolving networks. While early remote work boosts perceived productivity and autonomy, sustained WFA may impede sensemaking and work-life boundaries. Over time, WFA will likely evolve into diverse configurations; optimizing career outcomes requires balancing remote/on-site time and cultivating knowledge networks that promote sharing while reducing hiding. The study shows that moderate RWT maximizes CD and that high KS combined with low KH strengthens the positive effects of remote work on careers. These insights inform the design of hybrid work and HRM practices in the digital economy.

• Context specificity: Data from seven manufacturing firms in Zhejiang Province, China; results may not generalize to other regions, sectors, or cultures. - Sector limitation: Manufacturing focus; future studies should test in other industries. - Conceptual nuance: KS may have dark sides and KH may have situational positives; future work should unpack these gray areas and boundary conditions. - Level and variables: Micro-level analysis; future cross-cutting approaches could incorporate social and psychological variables and multilevel designs. - Measurement: CD proxied by career control; development of more comprehensive CD measures is warranted.