透明度(行为)
任务(项目管理)
计算机科学
对偶(语法数字)
识解
人工智能
社会心理学
灵活性(工程)
人机交互
心理学
认知心理学
规范(哲学)
作者
Ping Liu,Ling Yuan,Zhenwu Jiang
标识
DOI:10.5465/amproc.2026.15572abstract
摘要
As algorithmic systems become increasingly embedded in task allocation, monitoring, feedback, and performance evaluation, algorithmic management (AM) often enhances execution reliability but may dampen innovative output. Yet existing research offers limited testable micro-level explanations for this asymmetry. To explain why AM tends to benefit execution more than creativity, we focus on employees’ task construal. Integrating the attention-based view (ABV) and construal level theory (CLT), we develop a moderated parallel-mediation model: perceived AM intensity increases concrete task construal (CTC; a “how-to-do” focus) while reducing abstract task construal (ATC; a “why-to-do” focus). These shifts generate a positive indirect effect of AM on task performance (TP) via CTC and a negative indirect effect of AM on employee creativity (EC) via ATC. Further, algorithmic transparency (AT) strengthens the AM?CTC link and attenuates the negative AM?ATC link, thereby amplifying the positive indirect effect of AM on TP via CTC and weakening the negative indirect effect of AM on EC via ATC. Using three-wave survey data from full-time employees in Chinese organizations (N = 339), we test the model with structural equation modeling and latent moderated structural equations (LMS). The results largely support our hypotheses. This study advances a process account of the execution–innovation tension under algorithmic governance by theorizing an “attention–task construal” mechanism. It also shows how transparency—an actionable design parameter—reconfigures the strength of AM’s double-edged effects.
科研通智能强力驱动
Strongly Powered by AbleSci AI