Modeling consensus emergence in groups using longitudinal multilevel methods
提出共识涌现模型（CEM），通过扩展标准多层次模型，研究群体内共享感知和情绪随时间动态变化的过程，并利用两个纵向工作单元案例验证其有效性。
Abstract Organizational researchers have long been interested in studying bottom‐up multilevel processes where lower level units (e.g., employees) in organizations interact to jointly create characteristics of higher level units (e.g., work groups). This article contributes to the literature on bottom‐up processes by detailing a statistical approach—the consensus emergence model (CEM)—that allows researchers to study emergence of shared perceptions and feelings or climates in groups over time. The described methodological approach extends standard multilevel methodology by examining residual variances within a growth model to account for dynamic change in group consensus. The CEM provides a formal test for consensus emergence. The approach also allows researchers to test explanatory models of consensus emergence by including person‐level, group‐level, and observation‐level predictors. We illustrate the CEM by applying the method to data from two longitudinal studies of work units. The first study investigated job satisfaction in military companies. Our second study examined professional archeologists working in groups on a field excavation mission and focused on fatigue at the end of the work day. Our analyses demonstrate the CEM's ability to detect and study emergence, and suggest that the CEM may be a valuable tool to help extend the study of emergence in organizational research.