How organizational support drives knowledge worker performance: evidence for dual motivational and cognitive resource pathways

Introduction Against the backdrop of the digital knowledge economy, insufficient motivation and sustained cognitive attention among knowledge workers relate to poor task performance. Drawing on social exchange theory and the Job Demands‑Resources (JD‑R) framework, this study investigated how two‑dimensional organizational support (instrumental work support and socio‑emotional value recognition) relates to Chinese knowledge workers’ job performance, with dedication (motivational resource) and concentration (cognitive attentional resource) as parallel mediators. Prior studies often adopt unidimensional perceived organizational support and undifferentiated work‑engagement constructs, overlooking separate motivational and cognitive pathways. Methods Survey data from 346 Chinese knowledge workers were analyzed. We revised the Utrecht Work Engagement Scale by removing the redundant vigor dimension to create a knowledge‑work‑oriented two‑factor engagement scale. SPSS and AMOS were used for statistical analyses including confirmatory factor analysis, correlation, structural equation modeling, and Bootstrap mediation tests, with demographic variables controlled. All data are cross‑sectional and cannot confirm causality. Results Both organizational‑support dimensions positively correlated with dedication, concentration, and job performance; work support yielded stronger cross‑sectional associations than value recognition. Dedication and concentration served as parallel partial correlational mediators, and concentration showed a marginally closer linkage with performance. Discussion This study adds non‑Western localized correlational evidence for the JD‑R model by distinguishing motivational and cognitive resource pathways. It offers differentiated practical suggestions for optimizing workplace resources to sustain knowledge workers’ engagement. Future longitudinal research is needed to test causal relationships given cross‑sectional limitations.