Algorithmic Control: A Legitimacy Perspective on Worker-level Implications (AlgoWork)

我们最初研究项目的主要成果是：(1)区分七种形式的算法控制（AC）的概念框架（5RSM框架）和测量感知AC的相应量表；(2)传统组织背景下交流管理的运用及其对员工影响的实证分析；以及(3)一个法理学网络，揭示了平台工作人员如何判断不同AC形式的合法性（例如，在自主性、公平性、隐私性、技术压力方面），以及这些评估如何反过来影响他们的认知反应（例如，继续他们的平台工作的意图）。到目前为止，我们的项目工作和结果表明，需要进一步的研究来全面理解空调系统对工人层面的影响（而不是个别空调形式的影响），以及更深入地了解工人对这些系统的行为抵抗反应（所谓的“算法行动主义”实践）（而不是他们对不同空调形式的主要认知反应）。这就产生了三个研究问题（RQs），它们代表了后续项目的重点：(1)交流系统的关键概念维度（和相关特征）是什么，这些维度/特征的典型组合可以在基于平台和/或传统组织中找到？(2)传统组织中使用的哪些AC系统配置被员工判断为（不）合法，为什么？(3)平台工作人员使用哪些算法实践来“反击”AC系统，从流程的角度来看，这些实践是如何相互联系的？为了解决RQ1，我们打算开发一个交流系统的分类法，并在此基础上，经验地推导出这种系统的原型。为了解决RQ2，我们将首先提取对传统工人合法性判断至关重要的AC系统的维度和特征，然后识别被工人判断为（不）合法的AC系统配置。为了解决RQ3问题，将开发平台上下文中算法活动实践的动态过程模型，建立在此类实践的分类之上。总体项目结果将为负责任的空调系统设计和使用提供可操作的见解和建议，并支持相应的监管工作。此外，该项目旨在为有关空调系统的使用及其对未来工作的影响的更广泛的社会讨论提供新的见解。

The main outcomes of our initial research project are (1) a conceptual framework (5RSM framework) that distinguishes seven forms of algorithmic control (AC) and a corresponding scale for measuring perceived AC; (2) empirical insights into the use of AC and its effects on workers in the context of traditional organizations; as well as (3) a nomological network that sheds light on how platform workers judge the legitimacy (e.g., in terms of autonomy, fairness, privacy, technostress) of different AC forms and how these assessments in turn affect their cognitive reactions (e.g., the intention to continue their platform work). Our project work and results to date have indicated that further research is needed to develop a holistic understanding regarding the worker-level implications of AC systems (as opposed to the implications of individual AC forms), as well as to gain a deeper understanding of workers’ behavioral resistance reactions (so-called “algoactivistic” practices) to such systems (as opposed to their predominantly cognitive reactions to distinct AC forms). This gives rise to three research questions (RQs), which represent the focus of the follow-up project: (1) What are key conceptual dimensions (and related characteristics) of AC systems and what archetypical combinations of these dimensions/characteristics can be found in platform-based and/or traditional organizations? (2) What AC system configurations used in traditional organizations are judged as (not) legitimate by workers, and why? (3) What algoactivistic practices do platform workers use to ‘push back’ against AC systems, and how are these practices connected to one another from a process perspective? To address RQ1, we intend to develop a taxonomy of AC systems and, based on this, empirically derive archetypes of such systems. To address RQ2, we will first extract the dimensions and characteristics of AC systems that are critical for legitimacy judgments by traditional workers, and then identify AC system configurations that are judged as (not) legitimate by workers. To address RQ3, a dynamic process model of algoactivistic practices in the platform context will be developed, building on a categorization of such practices. The overarching project results will provide actionable insights and recommendations for the ethically responsible design and use of AC systems and support corresponding regulatory efforts. In addition, the project aims to contribute new insights to the broader societal discussion regarding the use of AC systems and their implications for the future of work.